Vendors / Bar, brewery & taproom

GoTab

dossier live legacy rubric

Claims in scope
224
Scored
224
Assessed
222
Unknown
2
Not applicable
90
Values challenged
81

Identity

Owner
Private, independent (GoTab, Inc.). No parent company disclosed. Venture-backed; GoTab itself has not been acquired and is not publicly listed. GoTab is however an ACQUIRER: it bought opsi (back-of-house recipe management, live food costing, inventory, invoice processing and task lists) on 2025-03-26, announced on opsi's own site, whose footer now reads 'All rights reserved by GoTab, Inc.' opsi remains a separately-branded, separately-priced platform sold as GoTab add-on modules at +$45/mo each, and by its own founder's statement it stays 'a standalone platform' supporting integrations with other POS systems — so opsi capabilities are first-party to GoTab as a vendor but are not part of the base POS.
Founded
2016 (GoTab, Inc., Arlington/Rosslyn VA) — founding year not independently re-verified in this pass; treat as approximate
Scale
unknown — GoTab publishes no customer count, location count, ARR or GMV figure. Company is private and files nothing public. Its status page enumerates ~20 production components but no volume data.
Who it is for
Documented from the vendor's own integration directory and product set: breweries and taprooms (Beer30, Ekos, Ohanafy, Untappd, Draft Control, Live Tap Lists), self-pour tap walls (iPourIt, PourMyBeer, NAPA, Pourtek, Wine Emotion), food halls with multi-vendor revenue splits, hotels/resorts (MEWS, Infor, Stayntouch PMS), stadiums/entertainment and gaming venues (Intercard, ggCircuit, aboutGolf). Typically 1-20 locations, high-volume beverage-led and guest-led-ordering formats rather than table-service casual dining. Store-count distribution is not published.
Site
https://gotab.com/

Pricing

transparency: published · unit: mixed · processor lock-in: partial

Software
Published tiers, per month: Basic $15/mo, Pro $99/mo (marked most popular), Sync $229/mo. Explicit a-la-carte add-ons: additional counter or handheld POS +$25-$35/mo each; KDS +$25-$35/mo; kiosk +$25-$35/mo; Easy Tab +$25/mo (included Pro/Sync); 24/7 support +$35/mo (included Pro/Sync); accounting integration +$10/mo (included Pro/Sync); on-premise OPServer +$75/mo (included Sync); Loyalty + Memberships +$35/mo (included Sync); Fraud Protection +$35/mo (included Sync); Order Status SMS +$50/mo; Gift Cards +$35/mo; 3rd-party aggregation +$55/mo; multi-location management +$50/mo; automatic remittances $15/unit; opsi invoice/recipe/inventory/task management +$45/mo each. Among the most transparent published price sheets in restaurant POS — but the base tier is a shell and a realistic bar/brewery configuration stacks well past the headline $15.
Card processing
Published: 'starting at 2.40% + $0.15' per in-person transaction; +0.85% for American Express; +1.5% for international cards. Blended card-present only — no interchange-plus option published.
Contract
Contradictory across GoTab's own pages. https://gotab.com/pricing says GoTab 'does not require long-term contracts.' https://gotab.com/toast-vs-gotab says GoTab's contracts are '2-year standard terms, transparent, with flexible options available.' The merchant T&C resolves the mechanism: an Initial Term is specified on the Merchant Order Form, after which it 'will automatically renew until the last day of the current calendar month, and thereafter, will renew monthly at GoTab's current rates,' with either party able to give 14 days' written notice of non-renewal. Net: negotiated initial term (commonly 2 years per the vendor's own comparison page), then true month-to-month with a 14-day out.
Early termination
No software ETF or liquidated-damages clause found in the merchant T&C (termination is for material breach with a 30-day cure). But the Hardware Flex Leasing addendum accelerates: 'If Customer terminates the Agreement or this Addendum for any reason before the Hardware Balance is paid in full, Customer shall immediately pay GoTab the entire remaining unpaid Hardware Balance,' and 'Returned Equipment has no residual value to GoTab and does not reduce Customer's remaining financial obligation.' A de facto exit penalty for leased-hardware customers.

API posture

public API: open

Cost to integrate
unknown — no partner fee, revenue share or certification cost is published. GoTab runs a partner program, referral program and partner resource center, but commercial terms are not public. The price sheet lists no API access fee and gotab.com/toast-vs-gotab claims an 'Open API; native QuickBooks Online sync at no extra cost,' but no page affirmatively states API access is included at $0.
Webhooks
Not documented. GoTab's complete published documentation index (https://gotab.readme.io/llms.txt) contains guides for Client Credentials Grant, Authorization Code Grant and Postman testing, and reference sections for Authentication, Customers, Orders & Tabs, Tab Passes, Menus & Products, Modifiers, Inventory, Locations & Integrations, Users & Roles, Payments, Wallet and Labor — and no webhooks or event-subscription section, no rate-limit page and no published throttling quotas. The documented pattern is REST plus a GraphQL Explorer at gotab.io/api/graph. Scored unknown rather than no, since absence from an index is not an affirmative vendor statement.
Data export on exit
Partial and inferred. The API provides broad read/write coverage (List customers, Get a tab, List tab items, product/modifier/inventory reads), so a technically capable operator can extract most data on demand. No bulk CSV or warehouse export product is documented. THE POST-TERMINATION WINDOW IS JURISDICTION-SPLIT and the earlier flat statement that the merchant T&C specifies none was only true of the US document: the US terms indeed give no retrieval window (§11.3 terminates all licences immediately, §5.2 disclaims any backup obligation), but the AUSTRALIAN merchant terms at gotab.com/australia-merchant-terms-and-conditions grant one at clause 2.4 — GoTab 'will use commercially reasonable efforts to make available access to the GoTab Solution to enable Merchant to download reports and export extracts of Merchant Content', chargeable only where that access is given 'more than 60 days after the end of the Term'. So: no published window for a US operator, sixty free days for an Australian one. On ownership the T&C is unusually clear and favorable: merchant retains ownership of Merchant Content; GoTab owns Usage Data and may use Merchant Content only 'in an aggregated and anonymized form.'
Notes
METHODOLOGY WARNING ON THE COMPLAINTS SECTION: the complaints array is deliberately empty. G2, Capterra, TrustRadius, GetApp, Software Advice, SourceForge/Slashdot, Trustpilot and Reddit (both reddit.com and old.reddit.com, including the JSON search endpoint) all returned 403/404/blocked to automated fetching in this pass, and the session's web-search budget was exhausted before review mining. I found zero verbatim operator complaints I can source. Rather than invent plausible themes, I report none and flag the gap. GoTab's operator-review corpus is likely thin relative to Toast/Square in any case (small private vendor, niche install base), but that is an inference, not a finding. Re-run complaint mining with a browser-capable fetch path before treating the complaint set as empty in fact. STRUCTURAL NOTES: GoTab's developer docs are genuinely public and self-serve (no NDA, no sales call), with a documented sandbox and both OAuth 2.0 client-credentials and authorization-code grants — materially better than most quote-only competitors. A changelog exists and deprecated endpoints are marked in place ('Pay for a tab (Deprecated)'), though no formal deprecation-notice policy is published. GoTab is NOT on DoorDash's 2026 Preferred Integration Partner list (Checkmate, Chowly, Deliverect, Otter, PAR, Qu, Square, Stream, Toast, UrbanPiper), and DoorDash does not appear in GoTab's own integration directory at all — only Uber Eats, Deliveroo, Just Eat, Order with Google, Shipday and FIRST Delivery — even though gotab.com/faq mentions DoorDash in passing. Also notable: status.gotab.io publishes per-component status AND rolling 90-day uptime percentages (99.98%-100.0% across ~20 components, with third-party dependencies Stripe, Twilio, AWS, SendGrid, Heroku, GCP, Cloudflare and GitHub tracked separately). That puts GoTab in the very small group of restaurant POS vendors publishing rolling uptime numbers. No contractual uptime SLA with credits was found.

Capabilities

Every claim is binary and checkable. Grades: A primary documentation · B product documentation · C pricing or feature page · D marketing claim · E third-party reporting · F inference with no source. A yes on a differentiator claim requires A or B.

Order capture & FOH workflow

Yes

Provides a graphical floor-plan editor where an operator can draw sections, place tables of varying shape/capacity, save multiple floor-plan layouts per location, and assign sections to servers.

order-capture-floor-plan-editor

The researcher never consulted GoTab's actual product documentation at success.gotab.io. It carries 'How do I Create a Floor Map?' and 'What are Maps?', documenting that 'Maps only allows you to map out a floor plan of the spots created in zones. Once a map is created, it will now show on the POS as an additional option to find the spots when initiating a tab', plus a Map View POS default view and 'Creating a Dine In Zone'. A drag-and-place floor-plan editor over zone-defined spots is documented. https://success.gotab.io/knowledge/how-do-i-create-a-floor-map · retrieved 2026-08-01 adversarially verified

B
Yes

Order lines can be tagged to a seat number at entry, and the check can be split by seat afterward without re-keying items.

order-capture-seat-level

The KB article "How Do I Add Seats to Orders?" documents both halves of the claim. Tagging at entry: "Our seating feature allows you to designate items to specific seat numbers or the entire table, as well as to rename seats." and "Simply click one of the numbers above the item modifiers add a seat number to an item." Split without re-keying: "Enter "Split Pay" on the tab." then "Click "Split by Seat" in the upper right to easily split the tab by the designated seats." Seating is a location toggle (Location Settings--Full Service--POS Seating) and seats can be renamed from the Seat view. Searched all five surfaces via q210-grep across 877 docs ('seat', 'per.seat|seat number|by seat|seat-level'): the KB is the only place seats are documented, and the gotab.com hits are navigation chrome or Tripleseat/SevenRooms integration pages. Falsified if the split-by-seat control turns out to require re-entry rather than acting on the existing seat tags. https://success.gotab.io/knowledge/how-do-i-add-seat-numbers · retrieved 2026-09-05 adversarially verified

B
Partial

Items can be assigned to courses and held, with a fire-next-course action available from a handheld or terminal that releases only that course to the kitchen.

order-capture-coursing-hold-fire

Score stands but the stated basis is wrong and should be corrected: the note claims 'named course objects and fire-next-course not documented', yet product docs carry a dedicated 'Full Service: POS Coursing' article under POS Ordering, alongside 'rushing orders'. Coursing is a named, documented feature rather than an inference from a handheld rush/delay control; I did not read the article deeply enough to confirm an explicit hold-then-fire action, so partial holds. https://success.gotab.io/knowledge/pos · retrieved 2026-08-01 adversarially verified

B
Partial

Supports splitting a check by item, by seat, by even N-way, and by arbitrary dollar/percentage amount, and merging checks or tables back together, including after partial payment.

order-capture-split-merge

Guests share the bill with friends and split/tip on the shared tab. Split by item/seat/even-N and post-partial-payment merge not documented. https://gotab.com/products/mobile-ordering-payment · retrieved 2026-08-01

D
Yes

Opens a bar tab against a card pre-authorization with a configurable auth amount, supports incremental re-auth as the tab grows, and auto-closes stale tabs at end of day. differentiator

order-capture-bar-tab-preauth

Auth amount is configurable: Location Settings > Edit > Open Tab Settings sets a 'Pre-authorization Amount' (help-centre example $1) plus 'Preauth Initial Order', which instead authorizes the value of the first order. Re-auth: 'Incremental Auth on Send Order' 'reauthorizes every time an order is sent, helping to ensure that your guests are never able to add to an order with insufficient funds on their card', with scope None / Customer (guest QR orders) / POS (server-entered orders) / All, and a fund-source filter of all cards or credit-and-prepaid only, excluding debit. Stale tabs: 'if you forget to close the tab manually, our system will close the tab automatically at the end of the business day' and the card is charged a location-set default tip. https://success.gotab.io/knowledge/insufficient-funds-protection · retrieved 2026-08-06

B
Partial

An open check can be transferred between servers, tables, and devices, and the transfer is recorded in an audit log naming both employees.

order-capture-transfer-audit

Transfer is documented, the two-name audit log is not. "You can easily transfer any tab on the POS to another staff member", via More then "Press Transfer Tabs", choose the tab(s), and "Choose the user you would like to transfer the tab(s) to", after which "The newly assigned user will now be the owner of the tab(s) and the sales and tips will show on their user report". Manager-dashboard reassignment exists too (getting-started-tabs, Reassign) and table-level movement is covered by pos-server-spot-assignments. Shortfall: what is documented is post-transfer attribution to the receiving user only; there is no transfer record naming both the losing and the gaining employee, no timestamped transfer entry, and 'audit log|audit trail|activity log' returns only five of 877 docs, none about tabs. Searched all five surfaces. Falsified by any article or API field exposing a transfer history with the prior owner. This upgrades the standing unknown, whose own verifier note said the verdict was upheld only because the enum lacked an upgrade to partial. https://success.gotab.io/knowledge/how-to-transfer-tabs · retrieved 2026-09-05 adversarially verified

B
Yes

Ships a first-party handheld ordering app running natively on the handheld device (not a remote/mirrored desktop session) that can send to kitchen and take a card payment tableside.

order-capture-native-handheld

First-party handheld POS app on Samsung A17; takes orders, rush/delay to kitchen, and contactless payment tableside. https://gotab.com/products/pocket-phone-only-pos · retrieved 2026-08-01

C
Yes

With the internet down, terminals and handhelds can still create orders, apply modifiers, send to kitchen, and tender cash, and the vendor publishes what does and does not work offline.

order-capture-offline-order-entry

OPServer keeps POS, handhelds and kiosks creating orders and routing to printers/KDS offline; cash always available; vendor names the exclusions. https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-08-01

C
Yes

Guest QR at-table ordering writes items onto the same open check as server-entered items, rather than creating a separate order ticket. differentiator

order-capture-qr-same-check

'Just like any other tabs in GoTab, an Easy Tab is accessible from the POS and servers can also add orders to a guests tab. This is cleanly broken out on a guests tab allowing for easy identification of items they entered on their own versus items added by a server.' A server-initiated order authorizes the card, the guest is texted a link to that same tab and continues ordering on it; the guest/server split is a display grouping within one tab, not a second ticket. Guests can also share the tab so several phones order onto it. https://success.gotab.io/knowledge/how-to-use-easy-tab · retrieved 2026-08-06

B
Partial

Offers a first-party self-order kiosk that renders the same menu and modifier tree as the POS with no separate menu build, and the vendor documents ADA/accessibility conformance for it. differentiator

order-capture-kiosk-first-party

First-party kiosk shares the product catalog with no separate menu upload, but no ADA/WCAG conformance is published. https://gotab.com/products/self-ordering-kiosk · retrieved 2026-08-01

C
Partial

Supports order throttling: per-channel capacity limits per time slot with automatic quote-time extension when kitchen load exceeds a configured threshold. differentiator

order-capture-throttling

Documented Throttles feature the prior pass missed: 'Throttling is a functionality used to limit the number of orders that can be ordered by guests during a certain day and time', with an Order Limit and a Subtotal Limit per zone schedule step — 'if you want to limit orders to 5 orders every 10 minutes... Once the maximum orders have been reached for each 10-minute increment, guests will no longer be able to select that time to order.' Throttles are schedulable and can be toggled on/off. Shortfalls: full slots become unselectable rather than any automatic quote-time extension, limits are order-count/subtotal per time step rather than kitchen-load-driven, and 'Throttles can only be assigned to Takeout/Delivery zones.' https://success.gotab.io/knowledge/order-throttles · retrieved 2026-08-05

B
Yes

Accepts future-dated/scheduled orders with a per-channel lead time and automatically injects them into the make queue at a computed fire time rather than on receipt.

order-capture-scheduled-orders

Future-dated orders with computed fire-time display are documented end to end. Zone config carries 'Time needed to fulfill when the guest places an order', 'Furthest ahead takeout orders can be placed' and slot intervals of 5/10/15/30 minutes; the POS takeout flow selects order date and time. KDS settings gate injection: 'Show up to (Scheduled Orders): How much time do you want in advance for a scheduled order to display (i.e. 1 hour, 1 day, 5 days)', plus 'Hide Items Until Time to Prep' and 'Hide Scheduled Dine-In Orders'; the KDS Takeout doc adds that tickets 'show the time the takeout is scheduled for as well as when the ticket will be interactive', and with print-on-fulfill off a takeout order 'will fire once it is ready to be prepped' rather than on receipt. https://success.gotab.io/knowledge/kds-settings · retrieved 2026-08-05

B
Partial

Provides a distinct catering order flow with quotes, deposits, balance-due tracking, and a delivery/event schedule separate from the standard order queue.

order-capture-catering

Event Deposits are a documented advance-payment flow the prior pass missed: a 0%-tax open product plus a Non-sales Revenue (Deposit) processor takes the deposit ahead of the event, and on the day it is applied as a payment method to pay down the tab, with the remainder collected normally. Deposits can be invoiced by email with a payment-due date and paid online ('Event Deposits: How to Invoice a Guest'). Shortfalls: no quote/proposal generation, no separate catering order queue or event schedule (event management is positioned on the Tripleseat integration), and balance-due tracking is just the open tab remainder. https://success.gotab.io/knowledge/how-to-set-up-an-event-deposit · retrieved 2026-08-05

B
Unknown

Pushes an order-ready signal back to third-party marketplaces (e.g. DoorDash) from POS or KDS state, rather than requiring a tablet tap. differentiator

order-capture-order-ready-signal

CLAIM: pushes an order-ready signal to marketplaces from POS/KDS state. VERDICT: stays unknown. NEW SURFACE: the aggregator portal the old note called out-of-corpus is identified - gotab.streamorders.com is a Stream white-label tenant (Stream blog 2024-07-17: "Stream's latest white-label partner - GoTab!"), and Stream's help centre (help.streamorders.com, all 164 /en/ sitemap articles) was read. Closest evidence: "When you mark an order as 'Ready', a notification is sent to Deliveroo indicating that the order is prepared for collection ... Doordash, Uber Eats & Grubhub all function the same way as Deliveroo", but the article's POS example is Toast ("Marking an order as 'Ready' on Toast") and no Stream article names GoTab. The GoTab tenant's JS bundle has a console status progression in_progress -> ready_for_pickup -> out_for_delivery -> completed, which is a dashboard action, not evidence that GoTab KDS dispatch propagates. GoTab's webhook catalogue (docs.gotab.io/concepts/webhooks/) has no ready/fulfilled event. Resolved by a GoTab or Stream article stating that a GoTab KDS dispatch marks the DSP order ready. Retrieved 2026-09-16. || 322nd: VERDICT: unknown (held). NEW SURFACE (2026-09-17): the GoTab-branded Stream console (gotab.streamorders.com, production Next.js app-router tenant) serves PUBLIC SOURCE MAPS: 38 of the 39 chunks on /login have a 200 .js.map with sourcesContent (polyfills is the one 404), 377 sources outside node_modules, part of them Next.js internals (app files include utils/platform.utils.ts, utils/order.utils.ts, lib/atoms/mutations.ts, lib/stream-api.ts ...). Other routes' chunks are not enumerable (app router, no _buildManifest), so this is NOT a census of the console. utils/platform.utils.ts carries the per-integration capability table, and GoTab (POS.gotab) is in exactly two lists: DISPATCH_SETTING_INGEST_SWITCH [toast, square, example_pos, gotab] and MENU_EDITING_DISABLED_SOURCES [toast, zonal, akru, orange, // @TODO: remove below here and source from pos config correctly, gotab, shift4]. DISPATCH_SETTING_INGEST_SWITCH has no comment and no other reference in the mapped sources, so its meaning (a dispatch/courier setting switch during POS ingest) is unstated and it is NOT evidence that a GoTab KDS/POS state pushes "ready" to DSPs. utils/order.utils.ts defines the Stream order status READY_FOR_PICKUP ("Order is ready to be picked up by the customer, or delivery driver") and the console's manual next-status map (in_progress -> ready_for_pickup -> out_for_delivery), a console action. Controls over those sources: dispatch_setting 0, order_ready 0, pos_status 0, status_sync 0, fulfillment_status 0. Resolved only by a GoTab or Stream statement that GoTab fulfilment state propagates to the DSP. Retrieved 2026-09-17. || 330th: 330th, NEW BOUND, cell unchanged. The success.gotab.io KB was re-derived to an HREF FIXPOINT rather than inherited: sitemap 278 URLs + 26 category/hub pages reachable only from article bodies = 304 pages, 825,520 bytes of extracted text, one crawl round to zero new hrefs. Two enumerations in it bear on this claim. (1) 'KDS: Ticket Functionality' enumerates the COMPLETE action set behind the ticket's [...] menu: 'Text & Fulfill', 'Text - Order Ready', 'Text see staff', 'Custom Chat', 'Prep All Items', 'Print To', 'Reset Order', 'Rush', 'Batch Orders'. Every messaging action targets the GUEST, and the article states the precondition: 'In order to use the texting functionality, the guest must place an order on their phone' - which a DSP-injected ticket cannot satisfy. 'Text - Order Ready: This sends the guest a text to let them know their order is ready. Keep in mind this does not fulfill the order on the KDS.' There is no action, and no auto-fulfilment setting ('Auto-Text on Fulfillment' also texts the guest), that targets the marketplace. (2) 'How To Setup 3rd Party Ordering' describes the whole flow and stops at ingest: 'your DoorDash, UberEats etc. orders are facilitated through the 3rd Party Ordering platform into GoTab. From GoTab, you can view the orders in your POS and fulfill the orders on your KDS.' - nothing about status flowing back out. Tokenised controls over the 304 pages: 'order ready' 2 (both the guest SMS above), 'dispatch' 0 in a DSP sense. ALSO RE-CENSUSED TODAY: Stream's help centre (help.streamorders.com/sitemap.xml, 204 URLs, 167 EN articles - 3 more than the 164 held from the 317th: slice-integration-faq, how-to-connect-hungry-panda, how-do-i-get-started-with-stream-eu, all fetched) still contains ZERO occurrences of 'gotab' in any article. Held unknown, not `no`: the outbound leg would be Stream's partner integration to the DSPs, which is not a published surface, so no census of GoTab's own docs can bound it. Retrieved 2026-09-18. adversarially verified

F
Yes

Voids, comps, and discounts require a reason code and can be gated behind role-based manager approval, with all instances recorded in an exception report.

order-capture-void-comp-controls

Documented, not just inferred from the User Roles API. The comp/void article instructs 'Press comp or void at the top of the tab' then 'Select a reason' — reason codes exist. The Service (Employee Reports) page confirms manager gating and attribution: comps, voids and discounts each 'list out the employee that initiated' the action plus 'the manager that PIN'd in to approve'. Reason codes, manager approval and an exception report are all present. https://success.gotab.io/knowledge/how-to-comp-or-void-items · retrieved 2026-08-01 adversarially verified

B

Menu, modifiers & pricing engine

Partial

Supports modifier groups nested at least three levels deep, each level with independent min/max selection counts and a forced-selection (required) flag.

menu-pricing-nested-modifiers

Modifier CRUD plus Item Groups and Update Item Group Selection in the public API imply selection constraints; nesting depth and min/max/forced flags not documented. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
No

A single modifier can carry different prices depending on the parent item and the selected size, configured as a matrix rather than by duplicating the modifier.

menu-pricing-modifier-price-by-parent-size

The modifier-creation doc enumerates the complete field set, and price is a single fixed field: 'Price: This amount will be added to the item's base price.' The option-group doc likewise prices each option as one flat amount, and per-variant pricing exists only on product variants (Variant Details), not on modifiers attached to them. No price matrix by parent item or parent size appears anywhere in the catalog tooling (creatingamodifier, intro-to-option-groups, menu-options). https://success.gotab.io/knowledge/creatingamodifier · retrieved 2026-08-04

B
No

Supports modifier quantity tiers (light / regular / extra / double) with a configurable price multiplier per tier, distinct from adding the modifier twice.

menu-pricing-topping-quantity-tiers

GoTab's documented mechanism for multiple amounts of a modifier is Option Quantities — a min/max on how many selections a guest may make ('Use option quantities if you want to allow your guests to select multiple selections of your modifiers!'), which is exactly the add-it-N-times pattern the claim distinguishes itself from. No light/regular/extra tier object and no per-tier price multiplier exists in the modifier docs; creatingamodifier lists the full field set (Name, Short Name, Description, Open Text Mod, Add Product As Option, Price, Allow Multiple Selections, Require) with a flat price only. https://success.gotab.io/knowledge/option-quantities · retrieved 2026-08-04

B
Partial

Upsell/suggestive-sell prompts can be configured per item and per channel (POS, kiosk, online), and the vendor reports attach-rate on those prompts. differentiator

menu-pricing-upsell-prompts

Kiosk page claims increased check sizes and upsells; per-item/per-channel prompt configuration and attach-rate reporting not documented. https://gotab.com/products/self-ordering-kiosk · retrieved 2026-08-01

D
Partial

A single 86 action propagates automatically to POS, KDS, the vendor's own online ordering, kiosk, and connected third-party marketplaces, with a documented propagation latency.

menu-pricing-86-propagation

Operators can 86 menu items or update inventory levels right from the line, and 86ing works offline; propagation to third-party marketplaces and any latency figure not documented. https://gotab.com/features/online-ordering · retrieved 2026-08-01

D
Yes

Supports per-item countdown/par counts that decrement on sale and automatically 86 the item at zero, with a scheduled auto-restore. differentiator

menu-pricing-countdown-auto-86

Tracked stock levels decrement with orders (Reserved vs Available quantities, depletion reports) and carry a per-product out-of-stock action: '86 Product: When stock levels hit 0, the product will be temporarily disabled and automatically re-enabled the following day', shown as out of stock on the guest-facing menu; 'Disable Product' alternatively removes it until manually re-enabled. The auto-restore is a documented toggle: Location Settings > Edit > Re-enable Unavailable Products Overnight ON/OFF (how-to-86-or-disable-an-item). Par levels flag low stock in red. https://success.gotab.io/knowledge/inventory-management · retrieved 2026-08-05

B
Partial

Menus, individual items, and prices can be scheduled to activate and deactivate automatically by day and time in the location's own timezone.

menu-pricing-dayparting

Set up daily, seasonal and special event menus; automatic activation windows and location-timezone handling not documented. https://gotab.com/features/online-ordering · retrieved 2026-08-01

D
Yes

Supports distinct price books per channel (dine-in, pickup, delivery, kiosk, each third-party marketplace) including percentage-markup rules applied to a base price rather than manual per-item entry.

menu-pricing-channel-price-books

Confirmed verbatim: 'Edit dollar amount, percentage markup per category and item or store for each marketplace within seconds', with the GoTab catalog flowing to third-party marketplaces. Per-channel price books are real. Note the page names no marketplace by name, which supports the dossier's separate partial on 3P coverage. https://gotab.com/features/online-order-aggregation · retrieved 2026-08-01 adversarially verified

C
No

Supports dual pricing / cash-discount at the item level with the card price displayed as the base price, applied consistently across POS, kiosk, and online ordering, and printed on the receipt. differentiator

menu-pricing-dual-pricing

GoTab's documented card-cost-recovery mechanism is the opposite model: a percentage surcharge (max 3.0%, 2.0% in Colorado) added to credit transactions at payment, itemized on check and receipt, with debit/prepaid/gift excluded — a fee on top of a single posted price, not two stored prices with the card price displayed as base. The help centre documents no cash-discount or dual-pricing mode, and GoTab's own explainer (gotab.com/latest/payment-processing-costs-explained) walks through surcharging, dual pricing and cash discounting as strategies while the surcharging feature is the only one that exists as documented product. https://success.gotab.io/knowledge/credit-surcharging · retrieved 2026-08-04

B
Partial

Menu items and modifiers can be linked to recipes/bills of materials so that a sale depletes theoretical inventory and yields item-level theoretical food cost. differentiator

menu-pricing-recipe-linkage

Delivered by partners — opsi recipe management billed at +$45/mo on GoTab's own price sheet, plus meez, Craftable, Smart Bar USA. Not native. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

Pushes menus to third-party marketplaces through a certified direct integration (not scraping or a menu-management middleman) and surfaces per-item sync status and rejection errors to the operator.

menu-pricing-3p-menu-push

GoTab catalog and menu items flow to third-party marketplaces with per-category/item markup; per-item sync status and rejection errors not documented, and DoorDash/Grubhub are absent from the partner directory. https://gotab.com/features/online-order-aggregation · retrieved 2026-08-01

C

Payments & money movement

Partial

The POS operates with third-party payment processors of the merchant's choosing (documented list of supported processors/gateways) rather than requiring the vendor's in-house processing. differentiator

payments-processor-choice

Merchant T&C names Stripe, Paysafe and Adyen, so there is a documented choice set — but the FAQ says GoTab processes the payments and deposits the funds. Not bring-your-own-acquirer. https://gotab.com/merchant-terms-and-conditions · retrieved 2026-08-01

B
Partial

Vendor publishes card-processing pricing publicly (a specific flat rate or interchange-plus markup) rather than quote-only. differentiator

payments-published-rates

Published: 'Starting at 2.40% + $0.15' for card-present, with the footnote '*Rates quoted are for a location doing less than $5m/year in processing for Mastercard, Visa, and Discover. Add 0.85% for American Express, JCP, Corp, and 1.5% for International (Intl) transactions.' Shortfall: this is a floor rate on a first-party sales page, not a committed schedule - no interchange-plus markup is disclosed, no card-not-present or keyed rate is given, and the Merchant Terms and Conditions place Fees on the per-merchant Order Form. No rate card exists anywhere in the help centre (success.gotab.io) or the developer documentation (docs.gotab.io). https://gotab.com/pricing · retrieved 2026-08-06

C
No

Native dual-pricing / cash-discount mode stores two prices per item and prints both cash and card totals on the guest check and receipt. differentiator

payments-dual-pricing

No two-prices-per-item mode exists. GoTab's documented card-cost recovery is the opposite mechanism — a credit surcharge (max 3.0%, 2.0% in Colorado, debit/prepaid/gift excluded) added at payment atop a single posted price, and catalog docs price products and modifiers with one flat Price field. The closest documented feature is Cash Tax-Inclusive, a per-display toggle that folds tax into the cash price ('build the tax amount into the purchase price of a product paid with cash'); it stores no second price per item, and the dual regular/credit totals appear only on the CFD under surcharging. Matches the verified 'no' on menu-pricing-dual-pricing. https://success.gotab.io/knowledge/credit-surcharging · retrieved 2026-08-05

B
Yes

Surcharging engine automatically excludes debit and prepaid cards via BIN/product-code detection, enforces the network percentage cap, and can be enabled or disabled per location. differentiator

payments-surcharge-guardrails

The program details document each guardrail: 'our surcharging feature does not apply surcharges to debit, prepaid, or gift card transactions (any method other than credit cards)' — automatic card-type exclusion; 'the maximum percentage allowable to be passed along to customers is 3.0%' with 'Merchants in Colorado can charge a maximum Credit Card Surcharge of 2.0%' and CT/ME/MA/OK merchants ineligible — cap and jurisdiction enforcement; adoption is per-merchant opt-in ('If you choose to implement GoTab's surcharging feature... For accounts with surcharging in place'), with disclosure auto-surfaced 'on guest checks, receipts and Customer Facing Displays'. GoTab also 'may disable any fees or service charges... intended to circumvent proper credit card surcharging'. BIN/product-code detection is not named as the mechanism, but the automatic debit/prepaid exclusion it exists to guarantee is documented. https://success.gotab.io/knowledge/credit-surcharging · retrieved 2026-08-05

B
Yes

All first-party payment terminals accept EMV chip and NFC contactless including Apple Pay and Google Pay.

payments-emv-nfc

Adyen payment terminals are the first-party card devices (covered under GoTab's warranty); Apple Pay, Google Wallet and contactless are named across kiosk and handheld pages. https://gotab.com/gotab-limited-warranty · retrieved 2026-08-01

B
Partial

Accepts contactless payments directly on a phone (Tap to Pay on iPhone and/or Android tap-to-pay) with no separate card reader. differentiator

payments-softpos-tap-to-pay

Handheld (Samsung A17) accepts contactless digital-wallet payments; GoTab also says pair it with a separate terminal for chip and swipe. Tap to Pay on iPhone / Android tap-to-pay not named as the mechanism. https://gotab.com/products/pocket-phone-only-pos · retrieved 2026-08-01

D
Yes

Vendor ships a handheld that takes EMV/NFC payment at the table, including tip prompt and check splitting, without carrying the card away.

payments-pay-at-table

Handheld takes orders and payments tableside; Easy Tab dips/taps the guest card at the table to open the tab. https://gotab.com/products/pocket-phone-only-pos · retrieved 2026-08-01

C
Yes

Guest can scan a QR code on the printed check or table tent to view and pay the check from their own phone, with the payment closing the check in the POS automatically. differentiator

payments-qr-guest-pay

Guest flow, from the help centre: 'Scan the QR code at your table', browse the menu, 'Select View Cart ... to begin the checkout process', 'Confirm that the order is correct then pay', then view the receipt. On the review-tab screen the guest gets three preset tip percentages plus a custom amount and 'the option to choose a payment method before moving on to close the tab'. The tab being closed is the same tab the POS holds - the Easy Tab article states an Easy Tab 'is accessible from the POS and servers can also add orders to a guests tab'. If the guest never closes it, 'our system will close the tab automatically at the end of the business day'. https://success.gotab.io/knowledge/open-a-tab · retrieved 2026-08-06

B
Yes

Supports both pre-auth-then-tip-adjust and on-device tip prompt flows, with a documented batch/adjust window and a manager screen for unadjusted tips.

payments-tip-adjust

All four claim elements are documented. Auth-then-capture is the platform model (a tab opens on a preauthorization; 'when your check is finalized, the system captures the actual amount'). Adjust window: 'You can easily add a tip to any closed tab for up to 72 hours'; previous-day tabs are reopened from the Manager Dashboard tabs page then tipped under All Tabs--Prev Unpaid on the POS, with tips reported on the day the tip payment is captured. Unadjusted-tip screen: POS 'all closed' tabs > 'View Tabs With No Tip', which lists tabs and adds tips (enter the total and 'GoTab will do the math for you'). On-device prompts: QR checkout presents three preset percentages plus a custom tip, with a configurable tip-suggestion calculation. https://success.gotab.io/knowledge/how-to-add-a-tip-to-a-closed-tab · retrieved 2026-08-05

B
Partial

Configurable tip pooling and distribution rules (by hours worked, sales, role percentage, or points) that produce a per-employee tip allocation exportable to payroll. differentiator

payments-tip-pooling

Delivered via TipHaus and eTip integrations rather than natively. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

Card payments can be captured while the internet is down via store-and-forward, with configurable per-transaction and cumulative offline limits. differentiator

payments-offline-store-and-forward

Same single sentence, same defects as reliability-offline-card-auth: vendor marketing page only, gated behind the OPServer add-on that Basic cannot buy, mechanism attributable to Adyen terminal store-and-forward rather than GoTab, and the dossier itself concedes no caps are published. Cannot stand as 'yes'. https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-08-01 adversarially verified

C
No

Vendor publicly documents who bears the loss when a stored offline transaction declines on reconnect, and surfaces a post-reconnect report of failed offline payments. differentiator

payments-offline-decline-liability

Neither half of the claim is published. The Payments report is the vendor's own enumeration of post-hoc payment views - "From the Payment Page in your manager dashboard, you can view payments, refunds, chargebacks and failures" and "you can search search by card number across a date range and filter by success, refunds, chargebacks and failures" - so 'failures' is a general failed-payment filter, not a post-reconnect report of stored offline transactions that declined. On liability, 'declin' appears in only 4 of 877 documents (two readme payment-terminal references where PAYMENT_DECLINED is a live checkout status, and two opsi.io legal pages that are not GoTab's); the Merchant Terms and Conditions carry no chargeback or offline-transaction risk allocation at all, Section 8 being a generic limitation of liability; and the OPServer offline product page (gotab.com/products/on-premise-server-with-enhanced-offline-mode) enumerates what does and does not work offline without ever mentioning declines, caps or who absorbs the loss. Searched all five surfaces. Falsified by a merchant terms exhibit or KB article stating offline decline responsibility. https://success.gotab.io/knowledge/payments · retrieved 2026-09-05 adversarially verified

B
Yes

First-party gift cards with real-time balance tracking, redeemable across every location in the group and through the online-ordering channel.

payments-gift-cards

Native gift cards as a $35/mo add-on; customer engagement page describes digital and physical gift cards. https://gotab.com/pricing · retrieved 2026-08-01

C
Partial

House/on-account tabs with per-account credit limits, running balance, and periodic statement or invoice generation.

payments-house-accounts

House accounts are a documented internal payment method: a House Account processor holds accounts of two types (User accounts follow selected users on every display; Display accounts bind to specific displays), each created with a starting balance, gated by the control:payments permission, and usable as a tender at Pay; account balances surface as liabilities on the payments dashboard ('accounts gives operators greater insight into account (house, gift card) transactions and liabilities, which are current balances'). Shortfalls: no per-account credit limit and no periodic statement or invoice generation is documented in any house-account article. https://success.gotab.io/knowledge/add-house-account · retrieved 2026-08-05

B
Partial

A single check can be settled with multiple tenders and split by seat, by item, by even shares, or by arbitrary amount, with no hard cap below eight ways.

payments-split-tender

Multiple guests pay portions of one shared tab and multiple tender types are accepted; N-way limits, by-item and by-seat splits not documented. https://gotab.com/products/mobile-ordering-payment · retrieved 2026-08-01

D
Partial

Voids, refunds, discounts, and no-sales require role-based authorization (manager PIN or approval) and write to an immutable audit log identifying the approver.

payments-refund-void-controls

Role gating and approver identity are documented for three of the four action types. The Service report "will display sales, comps, voids, and discounts and can be viewed by all or grouped by an employee", and comps, voids and discounts each "list out the employee that initiated" the action, plus "the manager that PIN’d in to approve" it. Refunds are permission-gated rather than PIN-approved (why-dont-i-see-the-refund-option-on-my-pos names the manage:ordering and manage:refunds permissions), refund reasons are configurable (how-do-i-create-refund-reasons-comp/void), and the payments dashboard shows who issued a refund and why (getting-started-viewing-customer-payment-information). Shortfalls holding this to partial: no-sales are the Paid In / Paid Out flow, gated only by the control:payments permission with no approver named; refunds have no documented second-party approval step; and nothing describes the record as immutable or tamper-evident - 'audit log|audit trail|activity log' hits five of 877 docs, none about payments. Searched all five surfaces. Falsified by a documented manager-approval prompt on refunds plus an append-only event log. https://success.gotab.io/knowledge/service-employee-reports · retrieved 2026-09-05 adversarially verified

B
No

In-product dispute/chargeback dashboard that lists open disputes and allows evidence submission assembled from the POS transaction record. differentiator

payments-chargeback-tooling

Downgraded to no. GoTab's own FAQ answers this cell in the negative: "Chargebacks can only be disputed through the Adyen balance platform which GoTab customers do not have access to." There is therefore no in-product dispute workspace and no operator evidence upload; the operator's only channel is the support mailbox - "We will email you via the support email listed on your manager dashboard" and "Submit additional evidence within 48 hours" - which is precisely the email chase this cell exists to detect. What GoTab does provide is automated representment by a partner that "includes all order and tab data associated with a chargeback from both the GoTab dashboard and the Adyen balance platform", plus a chargebacks filter on the Payments report (success.gotab.io/knowledge/payments); neither carries per-case state or an upload control, so the cell's two conjuncts both fail. Searched all four fetched hosts. Overturned by a disputes screen with case status and file upload. https://success.gotab.io/knowledge/chargeback-dispute-process · retrieved 2026-09-05 adversarially verified

B
Partial

Vendor publishes its deposit schedule and offers a next-day or same-day/instant funding option. differentiator

payments-payout-timing

Daily settlement is stated in the multi-vendor FAQ: 'Each vendor receives one consolidated deposit per day, representing their net sales after fees and any applicable deductions. Daily deposits keep cash flow consistent', with Adyen processing 'in real time' and depositing 'directly into each vendor's merchant account'; the Payouts dashboard documents deposits batched per fiscal day with a Settlement/Batch Date (accrual vs deposit views). Shortfalls: the daily cadence is published only in the food-hall FAQ context, no general deposit-lag schedule is published for standard locations, and no same-day/instant funding option exists. https://gotab.com/faq · retrieved 2026-08-05

C
Unknown

Vendor provides point-to-point encryption and a current PCI DSS 4.x Attestation of Compliance / P2PE listing available to merchants on request.

payments-p2pe-pci4

Nothing licenses a verdict either way, and the two halves of the claim fail for different reasons. Published attestation: the brief's control result shows /pci, /trust and /compliance all return the constant 66,059-byte gotab.com 404, so GoTab publishes no compliance page; 'P2PE' has zero hits in 877 documents and 'PCI' has eight, of which the only substantive GoTab statements are the marketing sentence repeated on gotab.com/faq and gotab.com/products/ipad-pos that the iPad POS uses the same PCI-compliant payment processing standards as traditional hardware (no version, no AoC) and an RFP-template guide telling buyers to ask about PCI DSS compliance. gotab.com/security offers only tokenization and end-to-end industry-standard encryption, which is not a PCI-validated P2PE solution and must not be read as one. On-request availability: the cell asks whether an AoC is available to merchants on request, and an AoC is normally issued under NDA rather than published, so a web census cannot measure it and the absence of a page is not the absence of the document. I refuse a yes (no P2PE listing, no 4.x version anywhere) and I refuse a no (on-request supply is unmeasurable from these surfaces). Resolved by a merchant trust portal, a sales-collateral AoC reference, or a named PCI-listed P2PE solution. ADDENDUM 2026-09-16: CLAIM: P2PE and a current PCI DSS 4.x AoC / P2PE listing available to merchants on request. VERDICT: stays unknown. NEW SURFACES: all 38 Partner Resource Center PDFs (hrefs to 7820027.fs1.hubspotusercontent-na1.net; 37 read via pdftotext). The PCI flyer (2022-10-17) says only "GoTab is proud to announce that we are a PCI DSS Level 1 Service Provider" and "(GoTab utilizes tokenization technology)" - no version, no AoC, no P2PE. Tokens p2pe / point-to-point / attestation / aoc: 0 of 37 PDFs. Visa GRSP (visa.com/splisting API, 8,870 listings) has no GoTab entry (control: Toast, Inc listed, PCI DSS to Mar 31, 2027). Refuse yes (no version, no P2PE listing); refuse no (an on-request AoC is unmeasurable by web census and the registry is not a census of validated providers). Retrieved 2026-09-16. || 319th: CLAIM: P2PE plus a current PCI DSS 4.x AoC on request. VERDICT: stays unknown. NEW SURFACE: crt.sh (%.gotab.io 92 names, %.gotab.com 10) gave 94 unique hosts; probed root, robots.txt and /swagger, /swagger/index.html, /swagger/v1/swagger.json, /api-docs, /openapi.json, /docs, /redoc, /graphql, /sitemap.xml, /llms.txt, /v3/api-docs on the live ones. No new anonymous documentation surface: api/beta/staging/pos/kds.gotab.io are the same 16.7 KB SPA shell (all spec paths 302 to /not_found); mcp2.gotab.io and datahub.gotab.io 401; ecomm-store-agent.gotab.io 401 "Missing required headers"; deprecated.docs.gotab.io, discover.gotab.io, es.gotab.io/com are Cloudflare 403 "CNAME Cross-User Banned" (refusal, not absence); sdk.gotab.io 404; docs.gotab.io/llms.txt lists 335 pages, identical to the q310 corpus except graphql-schema.txt and specs/gotab-api.json (both already censused). The one new first-party surface is shop.gotab.io, a Shopify hardware store whose products.json (limit=250; page 2 empty) is a complete 78-product catalogue. Across all 78 hardware listings (incl. Adyen S1F2, Adyen AMS1, Verifone P630, MagStripe Reader): p2pe 0, point-to-point 0; pci 2, both on one listing - Verifone P630 "PCI Validity PCI PTS 6", a device (PTS POI) approval, not a P2PE solution listing or a DSS AoC. No trust-centre host exists among the 94 CT hostnames (no trust./security./compliance. subdomain). Retrieved 2026-09-16. || 328th: 328th: the PUBLICATION half of this cell is now settled negative (see the reliability-pci-dss-4-attestation row: complete 1,653-page gotab.com census, P2PE 0, AoC 0, 4.0.1 0, and two newly-found merchant agreements with PCI 0 / P2PE 0 / attestation 0). NEW SURFACE WORTH NAMING: /merchant-terms-and-conditions and /australia-merchant-terms-and-conditions, ~110 KB each, are the first GoTab contract documents ever reached - prior passes recorded /terms, /msa, /legal and /dpa as returning the site's 404 body, and they still do; only these two slugs work. The Australian agreement's only 'failover' language is a payment-processing arrangement ('GoTab may also nominate an optional "failover" Payment Processor ... in the event of primary Payment Processor unavailability'), a homonym for the reliability cell and unrelated to P2PE. This cell nevertheless stays unknown rather than no, because its second conjunct asks whether an AoC is available to merchants ON REQUEST - an AoC is normally issued under NDA, so no web census can measure it. Retrieved 2026-09-18. || 330th: MERGE HOLD (proposed no/B, held unknown): The claim ends 'available to merchants on request', which no public register or contract read can bound, and the record holds GoTab's own 2022 statement that it is a PCI DSS Level 1 Service Provider: a `no` would contradict the vendor's only statement on the subject. The PCI SSC P2PE register read (6 listings, gotab 0, positive controls firing) settles the P2PE-LISTING limb only, and a processor's listed solution would appear under the processor's name. Not partial either: the flyer is 2022, pre-4.x, grade C. PROPOSAL AS WRITTEN: 330th: MOVED to `no` on the first conjunct, which is now measured against the authoritative census rather than a proxy. Prior passes checked the VISA Global Registry of Service Providers (which is a registry of service providers, not of P2PE validations) and refused a verdict. Today the PCI SSC's own P2PE listing was read in full via its public listing endpoint (POST https://www.pcisecuritystandards.org/wp-admin/admin-ajax.php with action=get_p2pe_solutions_data, the nonce taken from the public listing page; a GET-only page render returns an empty table). Denominators: 136 VALID P2PE Solutions, plus the Expired Solutions list, plus the Valid and Expired P2PE Applications and P2PE Components lists - six complete listings, 183 KB of JSON. Tokenised case-insensitive control `gotab`: 0 hits in all six. Positive controls in the same payload: Toast 2, Shift4 7, Adyen 6, Bluefin 99, Verifone 661. A 'PCI P2PE solution' exists only by appearing on this list - it is the program's own register, not a sample - so GoTab provides no P2PE solution, application or component, current or expired. This closes the conjunct that no web census could reach before, and it agrees with everything else on record: the complete 1,653-page gotab.com census has 'P2PE' 0 and 'point-to-point' 0, gotab.com/security offers only 'tokenization' and 'end-to-end' encryption (E2EE is not P2PE), the 37 Partner Resource Center PDFs have p2pe 0, and the 78-product shop.gotab.io hardware catalogue's only PCI token is the Verifone P630's 'PCI PTS 6' device approval. NEW AND ALSO NEGATIVE: the two merchant contracts found in the 328th were read end to end (gotab.com/merchant-terms-and-conditions, 42,990 chars; /australia-merchant-terms-and-conditions, 44,169 chars): PCI 0, P2PE 0, attestation 0, audit 0, encrypt 0 in BOTH - GoTab's own merchant agreement creates no compliance-evidence obligation of any kind, so there is not even a contractual right to an AoC on request. The second conjunct (an AoC supplied under NDA on request) remains unmeasurable by web census and the vendor's only statement about it is a 2022 flyer's 'GoTab is proud to announce that we are a PCI DSS Level 1 Service Provider' with no version; if the owner reads the claim disjunctively (P2PE OR an AoC), that sentence would support `partial`/C instead. Read conjunctively as written, the claim fails. Retrieved 2026-09-18. adversarially verified

F

Kitchen & production

Partial

Items route to specific KDS stations by item, category, order type, or revenue center, with routing rules configurable per location without vendor involvement.

kitchen-station-routing

I fetched the KDS product page in full. It documents Smart Prep Times, color-coding by fulfillment time, ticket timers, bump bar and guest messaging — and never mentions station or zone routing. The phrase 'zone-based routing' appears only in the string 'Multi-state item tracking, smart batching, zone-based routing' on gotab.com/toast-vs-gotab, which is a vendor-authored competitive sales asset, the weakest evidence class in this exercise. The dossier's own note already conceded operator configurability is undocumented. A yes resting solely on a Toast-comparison page is not a yes. https://gotab.com/products/kitchen-display-system-kds · retrieved 2026-08-01 adversarially verified

C
Partial

An expo screen consolidates items from multiple prep stations into a single order view and only marks the order complete when every contributing station has bumped its items.

kitchen-expo-consolidation

KDS consolidates counter, QR and online orders into one display with multi-item panels; the all-stations-must-bump completion rule is not documented. https://gotab.com/products/kitchen-display-system-kds · retrieved 2026-08-01

D
Partial

Items can be assigned to courses and held, then fired to the kitchen on demand from a server terminal, handheld, or expo screen. differentiator

kitchen-course-firing

Rush/delay from handheld and smart prep sequencing exist; explicit course assignment with a hold-then-fire action is not documented. https://gotab.com/products/pocket-phone-only-pos · retrieved 2026-08-01

D
Partial

Per-item cook times are configurable and the KDS staggers item start/display times so that all items on an order finish simultaneously. differentiator

kitchen-prep-time-pacing

Configurable per-item cook time: yes - the product catalog carries a 'Prep Time' field ('This will display to the guest how long it will take to prepare this item') with a 'Show Prep Time On Menu' toggle. Documented pacing off it: auto-timed coursing fires each course at '(Previous course hold time + prep time + eating time) - (Current course prep time)', with the article warning 'It is important that you have correct prep times configured for your items for this setting to work as intended'. Named shortfall: no per-item staggering inside a ticket. The KDS ticket timer in countdown mode takes 'the longest prep time' among the items on the ticket for the whole ticket, and the only prep-time-aware item handling on the KDS is the Location Settings 'KDS/Chit Sort Order' option 'By Prep Time: Ascending or Descending', which reorders the display rather than delaying item start times. https://success.gotab.io/knowledge/coursing-1 · retrieved 2026-08-06

B
Partial

KDS supports physical bump bars or programmable keypads in addition to touchscreen interaction, with documented supported models.

kitchen-bump-bar-hardware

Bump bar is named as a KDS feature; no supported hardware models are published. https://gotab.com/products/kitchen-display-system-kds · retrieved 2026-08-01

D
Partial

An all-day view aggregates outstanding quantities per item (and per modifier) across all open tickets on the station.

kitchen-all-day-counts

The Multi-Items Panel is a documented all-day view the prior pass missed: 'The panel shows multiples of the same items across all orders and all items on the screen', with a 'multi-items' view for duplicates and an 'all-day' view for 'all items ordered on the screen that need to be prepared even if it's just one item'; KDS settings can open it by default ('Show Multi Item/All Day Panel By Default'). Shortfall: the doc states 'the multi-item panel does not display item notes or modifiers', so per-modifier aggregation is absent. https://success.gotab.io/knowledge/kds-display-functionality · retrieved 2026-08-05

B
Partial

Configurable target cook times per station or order type, with visual color escalation and/or audible alert when a ticket exceeds target.

kitchen-sla-alerts

The KDS page supports only 'Orders are color-coded based on the time they should be fulfilled' and 'View ticket timers with all products and their respective prep times.' That is visual aging, not an SLA alert: no configurable threshold, no escalation, no notification, no breach reporting. The dossier's own note ('configurable per-station targets not detailed') describes a partial, not a yes. https://gotab.com/products/kitchen-display-system-kds · retrieved 2026-08-01 adversarially verified

C
No

If a KDS screen or kitchen printer goes offline, tickets automatically fail over to a designated backup printer or screen with no ticket loss. differentiator

kitchen-printer-fallback

The vendor's own remedy for a down kitchen printer is entirely manual: power-cycle the printer, verify its IP, trace and power-cycle the switch, then power-cycle the GoTab box; jobs sit as 'pending print jobs' until the printer returns. No automatic reroute to a backup printer or KDS is mentioned, the KDS settings doc enumerates the complete Devices & Printers configuration (Fulfillment Printer, Closeout Printer, Pair to TDS) with no backup or failover field, and the POS Notifications doc lists no printer-offline alert. https://success.gotab.io/knowledge/what-to-do-if-one-printer-is-down-but-other-printers-are-working · retrieved 2026-08-04

B
Partial

KDS continues to receive and display tickets from on-site POS terminals when the internet or cloud backend is unreachable. differentiator

kitchen-offline-operation

Documented: GoTops is a locally installed 'POS and KDS app', downloadable for iOS, Android, Windows and Mac - 'Never worry about how to use GoTab during an internet outage. You can easily download the GoTops App on any device.' Named shortfall: no documentation describes the KDS continuing to RECEIVE tickets from POS terminals during an outage. The 5/9/25 POS Order Entry Enhancements note describes the opposite for the base product: 'You will still see the offline notification banner if your POS loses network connectivity, but you will now be able to continue adding items to an in-progress order ... Once the POS has reestablished its network connection, you will then be able to send the order.' Continuous POS-to-KDS routing during an outage is asserted only by the OPServer marketing page; a help-centre search for 'OPServer' returns zero articles. https://success.gotab.io/knowledge/downloading-and-using-the-gotops-app · retrieved 2026-08-06

B
Partial

A bumped ticket can be recalled/unbumped, and an individual item can be refired or reprinted without re-entering the order.

kitchen-recall-refire

Refire/reprint without re-ringing is documented: 'Reset Order will unmark any item marked as prepared or sent' (without impacting the ticket timer) and 'Print To' redirects a chit to any printer; KDS search queries 'across the entire day and across both fulfilled and unfulfilled tickets', so completed tickets remain findable. Shortfall: no explicit unbump/recall action that returns a fulfilled ticket to the active queue is documented in the KDS ticket-functionality, settings or display-functionality articles. https://success.gotab.io/knowledge/kds-ticket-functionality · retrieved 2026-08-05

B
Partial

Edits made in the POS to an order already displayed in the kitchen visually flag the added, changed, or removed items on the live ticket. differentiator

kitchen-order-modification-alerts

Multi-state item tracking is claimed on the KDS; visual flagging of POS edits to a live ticket is not specifically documented. https://gotab.com/compare/toast-vs-gotab · retrieved 2026-08-01

D
Partial

Reports actual ticket and per-station times (average and percentile) sliced by daypart, station, and order channel, exportable to CSV or an API.

kitchen-speed-of-service-reporting

The KDS Fulfillment Report exists: 'Avg Time = The average processing time of ALL orders that have been sent to the KDS during the selected day (and the selected time range if specified)', plus 'the top 10 slowest fulfilled orders of the day' for each selected date. Shortfalls: average and slowest-10 only — no percentiles, no per-station or per-channel slicing, and no CSV/API export is documented on the report. https://success.gotab.io/knowledge/fulfillment-report · retrieved 2026-08-05

B

Delivery, dispatch & third-party channels

No

Product includes a native in-house driver module: drivers are first-class entities with clock-in/out, assignment state (in-store / on-run / returning), and per-driver run history reportable from the POS itself (not via a third-party add-on).

delivery-driver-roster

GoTab does ship a native self-delivery module, and this article enumerates exactly what it is: the Pack & Route dashboard allows you to 'generate labels and packing slips, mark orders as complete, and plan delivery routes', and routes are computed by scheduled time, 'the shortest distance for you drivers, or for the number of vehicles in your delivery fleet'. Drivers are a unit of routing capacity, not an entity: nothing in the six-article delivery set (pack-and-route, pack-route-route-planning, pack-route-packing-slips, how-to-access-your-delivery-routes, delivery-zone, how-to-create-a-delivery-map) assigns a route to a named driver, records a driver clock-in or run state, or reports per-driver run history, and the finished route hands off to Google Maps. Searched all four fetched hosts (success.gotab.io 275, gotab.readme.io 102, gotab.com 422, www.opsi.io 59): the term driver appears in 10 documents, none of them a driver-management doc, and the API reference has no driver object. Falsified by any KB article or API resource defining a driver role with assignment state or a per-driver run report. Residual doubt: the retired concierge driver app, which pack-route-route-planning still recommends to delivery drivers but whose own KB articles are now 404. https://success.gotab.io/knowledge/pack-and-route · retrieved 2026-09-05 adversarially verified

B
Yes

Delivery areas can be defined as arbitrary map polygons or by drive-time isochrones, not solely as a radius or a ZIP/postcode list. differentiator

delivery-zones-polygon

This is a flagged highest-bar cell and the researcher missed real evidence. Product docs cover 'Creating A Delivery Zone' and 'How To Create A Delivery Map': the operator draws the region in Google My Maps as a closed shape ('Type needs to be a single Polygon'), exports a .KML and uploads it to GoTab's Zones Dashboard. Genuine polygon delivery zones — though note the workflow is an external-tool KML import, not an in-product map editor, which is worth carrying in the note. https://success.gotab.io/knowledge/delivery-zone · retrieved 2026-08-01 adversarially verified

B
Yes

Each delivery zone carries its own independently configurable delivery fee, order minimum, and quoted promise time, applied automatically from the validated customer address.

delivery-zone-pricing

The delivery zone configuration exposes per-zone commercial rules: a 'Minimum amount of purchases (in dollars) required to place an order' is set on the zone, and a delivery fee is configured alongside it. Caveat for the note: the minimum is explicit in the doc, the per-zone fee is referenced rather than fully specified, so if the rubric is strict this is a strong partial rather than a clean yes. https://success.gotab.io/knowledge/delivery-zone · retrieved 2026-08-01 adversarially verified

B
Partial

Address entry is validated/geocoded against a mapping service at order time and rejects or flags out-of-zone addresses before the order is accepted.

delivery-address-validation

GoTab's guest ordering client forces the delivery address through Google Places at order time, and the shipped code makes the gate unambiguous. The address component is mounted inside the Maps loader as {libraries:["places"]} with the handler registered as [["ready","onGoogleMapsReady",false]]; onGoogleMapsReady builds a Places Autocomplete over the input id `autocomplete-address-input-${randomId}` with types:["address"] and setFields(["address_component"]), and its place_changed listener reduces the returned components through the map {street_number:"short_name",route:"long_name",locality:"long_name",sublocality_level_1:"long_name",neighborhood:"long_name",administrative_area_level_3:"long_name",administrative_area_level_1:"short_name",country:"short_name",postal_code:"short_name"}. Free typing is actively disqualified: a keyup handler on the same input sets `this.autocompleted=false`, and addCustomerAddress runs `if(t.autocompleted){...yield o.post(`/cust/addresses${...}`,{type:"JSON",body:{street_1,street_2,city,state,zip,country,type,to_customer}})}else{t.state.show_autocompleted_error=true}` — so a typed-but-unpicked address is never persisted and instead renders the error text 'Please select an address by typing in the field above.' clearAddress() resets autocompleted to false. The persisted fields are the Places components, not the raw string. The saved address then scopes availability server-side: getZoneOrderDays builds `/api/loc/${a}/zones/days` and appends `address_uuid=${n}` when an address is selected. SHORTFALL (the second half of the claim is not met): nothing establishes that an out-of-zone address is rejected or flagged before the order is accepted. Over all 170 served chunks (10.4 MB) there is no out-of-zone, not-deliverable, serviceable or service-area vocabulary, and no geocode, polygon or containsLocation logic — zero hits for each. Over GoTab's own 343-page / 952 KB docs corpus (docs.gotab.io/llms-full.txt) 'geocod' is zero, and the only near-misses are poison: 'outside patio area', 'validated as an email address', 'verify the ip address'. The published data models are affirmatively silent rather than merely unquoted: the GraphQL Address type carries street1/city/state/zip/country and NO latitude or longitude (zero coordinate fields in the whole 373 KB schema), and the Zone type carries requiresAddress:Boolean! but no geometry, boundary or delivery-map field at all. The operator-side KML polygon is therefore not evidence for this cell and is deliberately excluded from this note — it is already scored at delivery-zones-polygon, it describes operator-side geometry only, and its own article has the operator email the .KML to a GoTab Account Manager to upload. So GoTab validates and geocodes the address at order time (met, from shipped code) but no published surface shows an out-of-zone address being refused or flagged; the guest is at most steered by whichever days/times come back for that address_uuid, which is server behaviour nobody has observed. Falsified by any served string or KB article refusing an address outside the delivery map. Evidence is shipped code, not original sources: 0 of 170 chunks carry sourceMappingURL and .js.map returns 404, and the chunk hash rolls on deploy — though the gate, loader and error string are byte-identical in all three sibling hashes of this chunk held locally (1add80fd, ea3dfe34, fb8e2384). https://cdn.gotab.io/static/uuid_1-yi~42832916.ea3dfe34.js · retrieved 2026-09-15 adversarially verified

A
Partial

Can hand a first-party order to an on-demand courier marketplace (DoorDash Drive, Uber Direct, Nash, Relay or equivalent) natively from the POS, with the courier quote and status returning into the order record.

delivery-daas-dispatch

Shipday and FIRST Delivery are named integration partners; whether courier quote and status return into the GoTab order record is not documented. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

Offers direct, first-party certified API integrations with DoorDash, Uber Eats, and Grubhub marketplaces (vendor-owned, not solely resold through Otter/Deliverect/Chowly/ItsaCheckmate middleware). differentiator

delivery-3p-direct-integration

The standing note is falsified in its last clause. It says Grubhub is absent entirely, but this KB article handles Grubhub explicitly, with its own publishing cadence - "with Grubhub as the only exception. Grubhub publishes every Tuesday so once you connect Grubhub, it will not be live until the following Tuesday." - and gotab.com/pricing sells 3rd Party Aggregation at +$55/mo naming Doordash, Grubhub and UberEats together. All three marketplaces are therefore operationally supported and the integration directory is a marketing logo list, not a coverage census. The correction does not raise the verdict, though: this article opens "This integration allows you to publish your GoTab menus to your chosen 3rd party delivery partners via an integrated order aggregator" and routes the operator to a third-party portal at gotab.streamorders.com, which is the middleware-resold path the cell's parenthetical excludes from a direct, first-party certified integration. Partial: full DSP coverage, no evidence of vendor-owned certified APIs. Falsified by a GoTab-owned DoorDash or Uber Eats certified-integration listing. https://success.gotab.io/knowledge/3rd-party-ordering · retrieved 2026-09-05 adversarially verified

B
Partial

Third-party marketplace orders inject automatically into the POS as normal tickets that fire to KDS/printers without tablet re-keying or manual accept.

delivery-3p-injection

Online Order Aggregation is sold as a +$55/mo module and the FAQ confirms third-party orders require internet (they do not work offline), implying automatic injection; tablet-free auto-accept not explicitly documented. https://gotab.com/features/online-order-aggregation · retrieved 2026-08-01

D
Yes

Menu, pricing, modifiers, and photos publish from the POS master menu out to connected third-party marketplaces, including channel-specific price markups, without maintaining a separate menu in each portal.

delivery-menu-push

Catalog and menu items flow from GoTab to third-party marketplaces with per-category and per-item dollar or percentage markup. https://gotab.com/features/online-order-aggregation · retrieved 2026-08-01

C
Partial

Marking an item or modifier 86'd in the POS pushes item-level out-of-stock status to connected third-party channels in near real time, and restores it on availability (DoorDash requires item and item-option 86ing of new POS partners).

delivery-86-sync

86 status does propagate to DSP menus through the 3rd Party Ordering aggregator — the setup doc's own rejection FAQ confirms it: 'Why was a guest's order rejected? A: Most commonly, a product is disabled/86d in GoTab and resyncing is not yet complete across all platforms.' Shortfalls: that same sentence documents the sync as laggy (guests can still order during the resync window and the order is rejected rather than the item hidden), no latency figure or near-real-time commitment is published, and modifier/item-option 86 propagation to DSPs is not documented. https://success.gotab.io/knowledge/3rd-party-ordering · retrieved 2026-08-05

B
No

Store can be paused/deactivated on each connected third-party marketplace from inside the POS with a timed auto-reactivation, rather than only from each marketplace's own tablet or portal.

delivery-store-pause

The vendor's own FAQ rules out an in-POS pause and prescribes the external workaround: 'Q: Should I turn off my 3rd Party Zone to Pause Orders? A: No. This will lead to orders failing to push into GoTab but will not stop guests from ordering on the 3rd Party Platforms. You must pause any or all 3rd Party menus/orders in your 3rd Party Ordering Platform' — i.e. log into the gotab.streamorders.com aggregator portal. No pause control inside the POS or Manager Dashboard and no timed auto-reactivation is documented. https://success.gotab.io/knowledge/3rd-party-ordering · retrieved 2026-08-05

B
No

Produces a third-party reconciliation report matching marketplace payout deposits to POS-recorded 3P sales, itemizing commission, marketing fees, refunds/adjustments, and identifying missing or unpaid orders. differentiator

delivery-3p-reconciliation

The setup doc states the disqualifier outright: 'Payments and payouts for 3rd party ordering are handled 100% independently of GoTab.' Marketplace orders land with tender type API_Integration and no settlement data GoTab touches, so no report can match marketplace payout deposits, commissions or fee adjustments to POS-recorded 3P sales, and no reconciliation report exists in the Manager Dashboard reports set. https://success.gotab.io/knowledge/3rd-party-ordering · retrieved 2026-08-05

B
Partial

Vendor exposes integration health to the operator — failed/rejected order injections, connection status per channel, and alerts on failure — rather than failing silently. differentiator

delivery-injection-error-visibility

CLAIM: vendor exposes integration health to the operator - failed/rejected order injections, per-channel connection status, and alerts on failure. VERDICT: partial/D. Every conjunct is described, but only in the white-label PARTNER's documentation, and this is a differentiator, so the grade floor (A/B for yes) caps the value at partial. NEW SURFACE: GoTab's own 3rd Party Ordering doc (docs.gotab.io/operator/integrations/3rd-party-ordering/) sends operators to gotab.streamorders.com; that host serves a Next.js app titled 'Log In - GoTab', and Stream's own blog (www.streamorders.com/blog/gotab-selects-stream-for-their-third-party-order-and-menu-management-needs, July 17 2024) says "We're excited to introduce Stream's latest white-label partner - GoTab!". So the GoTab aggregator console IS a Stream tenant, and Stream's Intercom help centre (help.streamorders.com; sitemap 201 URLs, all 164 /en/ articles fetched = census of that sitemap) documents it. EVIDENCE: understanding-alerts: "1. Click on Alert Bell This will tell you how many unread notifications you have ... 3. Click on Show more Show more will give you details including why the order failed, when the order failed, and more! 4. Click on Failure Reason: Failure Reason will help you diagnose the issue". help.streamorders.com/en/articles/10950304-editing-alerts-for-members: "Click the Alerts dropdown and choose which alerts you want this member to receive". help.streamorders.com/en/articles/9923711-understanding-order-errors: an error table (ORDER VALIDATION FAILED, ITEM NOT IN POS, PRICE MISMATCH, OUT OF STOCK ...) with a column "Merchant Facing Reason (what they get in the email)". help.streamorders.com/en/articles/11145293-what-do-the-status-icons-mean: per-DSP status icons "Published The DSP is connected and the menu has been successfully published", "Error Something went wrong during integration or menu publishing. Hover icon for failure details.", "Paused". Corroboration from the GoTab tenant's anonymous JS (gotab.streamorders.com/_next/static/chunks/3744-e5579e4438f6e4da.js and siblings): a 'notificationsUnreadOnly' setting and an order-status enum with FAILED / active_failed. SHORTFALL / WHY NOT YES: none of the 164 Stream articles names GoTab (they are written against Toast/Clover/Square), the source is a third party (D), and GoTab's own docs describe none of these screens. Retrieved 2026-09-16. https://help.streamorders.com/en/articles/12662559-understanding-alerts · retrieved 2026-09-16 adversarially verified

D
Partial

Guests receive a branded order status / driver tracking page or SMS link on the restaurant's own domain for first-party delivery orders, driven by real order and driver state.

delivery-tracking-page

Guests get live order updates on their tab and two-way SMS; Order Status SMS is a +$50/mo add-on. A branded driver-tracking page on the restaurant's own domain is not documented. https://gotab.com/pricing · retrieved 2026-08-01

C

Digital ordering & guest-facing channels

Yes

Vendor ships a first-party, commission-free web ordering site on the restaurant's own branding/domain that writes orders directly into the POS with no per-order marketplace commission.

digital-first-party-web

Commission-free ordering accepted from your website, social media channels, email or any other platform with branded menus, writing into GoTab directly. https://gotab.com/features/online-ordering · retrieved 2026-08-01

C
Yes

The web/app ordering menu is generated from the same POS menu record as in-store — one item edit propagates to all first-party digital channels without a separate digital menu build.

digital-menu-single-source

One product catalog drives kiosk, QR, POS and online — no separate systems to manage or menus to upload. https://gotab.com/products/self-ordering-kiosk · retrieved 2026-08-01

C
Yes

Guests can create an account with saved addresses, saved (tokenized) payment methods, and one-tap reorder of a previous order across first-party channels.

digital-account-saved-payment

Wallet API supports adding cards, listing/deleting payment methods and setting a primary; guests reorder throughout a visit from the persistent tab. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Partial

Digital checkout presents configurable or algorithmic upsell/cross-sell suggestions (rules-based add-ons or personalized recommendations from order history), with attach-rate reporting on those suggestions. differentiator

digital-upsell-engine

Upsell and reorder prompting is claimed (a customer cites ~3x reorder likelihood); configurable rules and attach-rate reporting not documented. https://gotab.com/products/mobile-ordering-payment · retrieved 2026-08-01

D
Yes

Supports future/scheduled orders with per-daypart throttling — capacity limits on orders or items per time slot that automatically close slots when the kitchen is saturated.

digital-scheduled-pacing

Scheduled ordering plus slot-capacity throttling are both documented. Takeout/Delivery zones take future orders with a lead time, an advance-order horizon and slot intervals of 5/10/15/30 minutes (delivery-zone fields 9-12), and Throttles cap each slot: 'set the order limit on the Throttle to 5 and set the time step to 10. Once the maximum orders have been reached for each 10-minute increment, guests will no longer be able to select that time to order' — automatic slot closure at capacity, with an additional Subtotal Limit per slot and schedulable, toggleable throttle windows. https://success.gotab.io/knowledge/order-throttles · retrieved 2026-08-05

B
Partial

Single digital ordering flow supports pickup, curbside (with arrival check-in), dine-in/QR table ordering, and delivery, with mode-specific fees and prep timing.

digital-fulfillment-modes

Dine-in QR, takeout and delivery described in one flow; curbside with arrival check-in and mode-specific fee/prep configuration not documented. https://gotab.com/features/online-ordering · retrieved 2026-08-01

D
Yes

Supports guest-facing QR scan-to-order and/or scan-to-pay at the table that opens or attaches to the existing POS check, including splitting and tipping.

digital-qr-table

Scan-to-order and scan-to-pay attach to the same open tab, with tipping and checkout from the guest phone. https://gotab.com/products/mobile-ordering-payment · retrieved 2026-08-01

C
Partial

Supports group ordering: a shareable link where multiple participants add their own items to one order under a per-person or total spend cap, with split or single payment.

digital-group-ordering

Multiple guests join and order on one shared tab and split the bill — genuinely group ordering — but per-person or total spend caps are not documented. https://gotab.com/products/mobile-ordering-payment · retrieved 2026-08-01

C
Partial

Offers a distinct catering ordering flow with separate catering menu/minimums, lead-time rules, quotes or proposals, deposits, and invoice/house-account or ACH payment terms. differentiator

digital-catering-portal

The pieces exist as generic mechanisms rather than a catering portal: zones carry per-zone order minimums, lead times and advance-order windows; event deposits take prepayment and are invoiced by email with a due date, payable online; Pack & Route documents packing slips, bulk order export and route planning explicitly for 'catered orders'; PMIX's scheduled filter is pitched as 'helpful for any takeout or catering scenario'. Shortfalls: no distinct catering ordering flow or storefront, no quote/proposal step, and ACH or house-account payment terms for catering invoices are not documented (event-deposit invoices are paid as ordinary online tabs). https://success.gotab.io/knowledge/how-to-set-up-an-event-deposit · retrieved 2026-08-05

B
Partial

Vendor offers AI voice phone ordering (native or a named certified partner) that answers inbound calls and injects the completed order into the POS without staff transcription. differentiator

digital-voice-ai-phone

Practical AI and Pulse AI appear in the integration directory; GoTab does not describe an AI phone-ordering flow on any product page. https://gotab.com/integration-partners · retrieved 2026-08-01

E
Yes

Supports Order with Google — the restaurant's direct ordering link is provisioned to the Google Business Profile as an ordering option (ideally set as 'Preferred by Business'), rather than leaving marketplace links as the only choice. differentiator

digital-google-order

Order with Google is a named integration partner in GoTab's directory. https://gotab.com/integration-partners · retrieved 2026-08-01

B
No

Supports Apple Business Connect ordering placement — an 'Order Food' custom action on the Apple Maps place card pointing at the vendor's first-party ordering flow.

digital-apple-business-connect

GoTab built and documents the exact Google analogue of this claim and nothing for Apple. The Google Redirect integration lets the Order Online button within Google 'redirect your guests directly to ordering through GoTab', configured in the Manager Dashboard Integrations page where you 'Set the integration type to Redirect' and pick a Takeout Spot, Delivery Spot or both. Against that, searching all four fetched hosts (success.gotab.io 275, gotab.readme.io 102, gotab.com 422, www.opsi.io 59) for apple business connect, apple maps and place card returns zero hits in all 877 documents; the only Apple references in the corpus are Apple Pay tender and iPad hardware. With the KB census closed at 275 published articles with no gated tier, and a first-party map-ordering integration that covers Google and not Apple, this is a measured absence rather than an unsearched one. Falsified by an Apple Business Connect entry on gotab.com/integration-partners or a KB article on registering an Apple Maps custom action; an operator self-registering an ABC action pointing at their GoTab menu URL would not make this a vendor capability. https://success.gotab.io/knowledge/how-to-setup-google-redirect · retrieved 2026-09-05 adversarially verified

B
Yes

Loyalty accrual and redemption work inside the first-party digital ordering flow using the same guest identity as in-store, not as a separate post-purchase enrollment.

digital-loyalty-attach

Members are automatically recognized, no separate cards to swipe or codes to enter; benefits and coupons apply on the fly during ordering and payment. https://gotab.com/features/memberships-and-subscriptions · retrieved 2026-08-01

C
Partial

Discounts, coupon codes, and loyalty offers are defined once and honored identically across POS, kiosk, and first-party digital channels, with channel eligibility controls.

digital-promo-parity

Single catalog with coupons and promotions applied automatically across surfaces; explicit per-channel eligibility controls not documented. https://gotab.com/features/customer-engagement · retrieved 2026-08-01

D
Yes

Vendor documents that the operator owns and can export first-party guest records (email, phone, consented marketing status, order history) in bulk, without a fee or vendor approval. differentiator

digital-guest-data-ownership

GoTab markets itself as an Open Data platform where your guest data is yours to use, and the merchant T&C confirms merchant ownership of Merchant Content with GoTab's use limited to aggregated/anonymized form; a List customers API exists for bulk retrieval. https://gotab.com/merchant-terms-and-conditions · retrieved 2026-08-01

B
Partial

Digital checkout uses vendor-hosted fields or a hosted payment page so card data never touches the restaurant's page, with the vendor publishing PCI DSS 4.0 compliance (including the client-side script-integrity requirements effective March 2025) and supporting 3DS.

digital-checkout-pci-sca

GoTab does not store the credit card number, only a secure encrypted token. No PCI DSS 4.0 statement, no script-integrity (6.4.3/11.6.1) mention, no 3DS documentation. https://gotab.com/security · retrieved 2026-08-01

B
Partial

Digital channels support the same service-fee / dual-pricing / card-surcharge configuration as the POS, with the required guest-facing disclosure and correct handling of jurisdictions and card brands where surcharging is prohibited. differentiator

digital-surcharge-transparency

The surcharge program is platform-level with automated guest disclosure: 'For accounts with surcharging in place, GoTab will surface notification of surcharges on guest checks, receipts and Customer Facing Displays', itemized on the printed receipt, with jurisdiction handling built in (CT/ME/MA/OK ineligible, Colorado capped at 2.0%) and debit/prepaid/gift cards excluded by card type. Shortfalls: the doc never explicitly shows the disclosure inside the QR/online checkout screen before payment (guest checks, receipts and the CFD are the named surfaces), and entrance/menu signage compliance is stated to be the merchant's responsibility. https://success.gotab.io/knowledge/credit-surcharging · retrieved 2026-08-05

B

Guest data, loyalty & marketing

Partial

Maintains a single guest profile that automatically merges transactions from in-store POS, first-party web/app ordering, and kiosk under one identity (phone or email match), with documented dedup/merge behavior.

guest-loyalty-unified-profile

One tab identity spans QR, server, kiosk and RFID, and a Customers API exists; documented dedup/merge behavior on phone or email match is not published. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
No

Orders arriving from third-party marketplaces (DoorDash, Uber Eats, Grubhub) attach to a native guest profile with usable guest identity rather than landing as anonymous tickets. differentiator

guest-loyalty-thirdparty-identity-attach

330th: MOVED to `no`. NEW SURFACE: the IMAGES inside the KB articles, downloaded at full resolution and read as images (prior passes read only the extracted text, which omits them). The screenshot in 'How To Setup 3rd Party Ordering' showing what a marketplace order looks like inside GoTab is a tab whose header reads 'Tab Created 07/30 09:14 PM', one line item 'UBER - KNOWLEDGE B. ... 3rd Party Pickup $22.50', and a footer 'Tab assigned to 3rd Party [Reassign]'. So the ticket is owned by a ZONE/pseudo-user called '3rd Party', not by a guest record: there is no customer attached, no contact handle, and the only identity present is a display string in the tab name. The article's own prose confirms the payment side: 'Under payments and payment types, they're listed as API_Integration' and 'Payments and payouts for 3rd party ordering are handled 100% independently of GoTab' (the Payment Types report screenshot shows the tender literally as 'API_INTEGRATION' alongside Visa/Amex/Mastercard). The identity is also proven UNUSABLE by a second article: 'KDS: Ticket Functionality' states 'In order to use the texting functionality, the guest must place an order on their phone', so the guest-messaging channel GoTab does have cannot reach a DSP guest at all. BOUND: the whole KB re-derived to an href fixpoint (304 pages, 825,520 chars of text, 1 round to 0 new links) - `guest profile` 0 hits, and the only customer-identity mechanism documented anywhere in it is the in-venue rewards check-in by phone number at the POS/CFD; plus the complete published OpenAPI (65 paths/90 operations) whose GET /integrations/customers returns 'Array of customers sandboxed to the integration' and whose order endpoints accept no customer_profile_uuid; plus the complete 1,653-page gotab.com census whose 38 'guest profile' hits all belong to reservation/CDP partners. CONTROLS/POISON: 'marketplace' matches 304 of 304 pages because 'GoTab Marketplace' (a multi-tenant venue feature, unrelated to DSPs) is in every page's nav sidebar - the real DSP vocabulary is '3rd Party Ordering', DoorDash 5 hits in 2 files, UberEats 6 in 2, Grubhub 4 in 2. Retrieved 2026-09-18. https://success.gotab.io/knowledge/3rd-party-ordering · retrieved 2026-09-18 adversarially verified

B
Yes

Supports at least two of points-per-dollar, visit/punch-count, and spend-tier accrual as configuration, without custom development or a bolt-on loyalty vendor.

guest-loyalty-accrual-models

Offers can key on specific item purchases, spend amounts or points assigned; MugClub-style club/visit programs are described. At least two accrual models as configuration. https://gotab.com/features/memberships-and-subscriptions · retrieved 2026-08-01

C
Partial

Supports status tiers with automatic promotion and demotion driven by rolling-window spend or visit counts. differentiator

guest-loyalty-tiers

Membership options with varying benefit levels exist in the catalog; automatic promotion/demotion on a rolling spend or visit window is not documented. https://gotab.com/features/memberships-and-subscriptions · retrieved 2026-08-01

D
No

Documents explicitly what loyalty lookup, accrual, and redemption do when the POS loses connectivity (queue-and-reconcile vs blocked). differentiator

guest-loyalty-offline-behavior

This is a documentation claim, and the documentation set that would have to answer it is present, exhaustive and silent. GoTab publishes an eleven-question offline block opening at 'What happens if the internet goes down? Will everything stop working?' and running through QR ordering, KDS, payments, open tabs, 'Will my inventory still update if I take orders offline?', 'Can I accept third-party delivery orders while offline?', menu edits, 'Can staff clock in and out if the internet is down?', mode indication and resync. Loyalty, memberships, coupons and gift cards appear in none of the eleven, and neither any KB article nor the OPServer product page addresses loyalty lookup, accrual or redemption offline. The GoTab Loyalty API reference classifies inquiry and redemption as customer-driven synchronous events posted out to an external URL, implying a hard connectivity dependency that GoTab nowhere states. Searched all four fetched hosts. Grade C because the instrument is a marketing FAQ rather than product docs. Falsified by any KB article specifying queue-and-reconcile or blocked behaviour for rewards check-in offline. https://gotab.com/faq · retrieved 2026-09-05 adversarially verified

C
Partial

Offers can be issued to a dynamically defined audience (rules over recency, frequency, spend, or items purchased) rather than only broadcast to the whole list. differentiator

guest-loyalty-targeted-offers

Customer segments are first-class in the API (add/delete customer from segment) and VIP-targeted coupons are described; rule-based dynamic audience definition over recency/frequency/spend is not documented. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Partial

Automatically computes recency/frequency/monetary or equivalent lifecycle segments (new, regular, at-risk, lapsed, VIP) without the operator building the queries. differentiator

guest-loyalty-rfm-segmentation

Not in the GoTab POS itself - GoTab Segments are manually built enrollment groups (unrestricted QR/link self-enrollment or restricted phone-number lists) with no behavioral computation. Behavior-derived segments come from Fishbowl, the guest relationship management platform GoTab acquired 2026-06-09 to let operators 'build richer guest profiles based on actual visit behavior' and 'segment audiences by frequency, spend, preferences, and loyalty status'; Fishbowl's GRM page documents 'Smart Segmentation - Create audiences in seconds with insights that reflect purchase patterns, dining frequency, and loyalty tiers' and that it 'uses machine learning to suggest guest segments'. Shortfalls: Fishbowl remains a separately branded product, the release describes the combined profile as arriving through 'deeper integration' rather than shipping today, GoTab explicitly keeps it optional ('or another best-of-breed solution'), pricing is not published, and no named lifecycle segments (new/at-risk/lapsed/VIP) are documented in either product's docs. The alternative documented path remains Klaviyo. https://gotab.com/latest/gotab-acquires-fishbowl-to-transform-guest-marketing-loyalty-and-repeat-revenue-for-hospitality-operators · retrieved 2026-08-06

D
Partial

Ships triggered lifecycle automations (birthday, first-visit thank-you, lapsed win-back) as configurable always-on campaigns rather than one-off manual sends.

guest-loyalty-lifecycle-automation

Delivered through Klaviyo, not natively: the documented integration streams 'real-time events related to guest behaviors' into Klaviyo, where private keys sync 'guest subscriber events... for marketing communications' and operators build lists, segments and flows. Shortfalls: birthday, first-visit and lapsed win-back automations must be configured and run in a separate Klaviyo account with its own subscription; GoTab's native marketing surface is coupons, discounts and manual segments with no documented automation engine or triggered campaigns. https://success.gotab.io/knowledge/connect-klaviyo · retrieved 2026-08-05

B
Partial

Sends both email and SMS campaigns natively from the platform, not solely by handing the list to a third-party ESP. differentiator

guest-loyalty-native-email-sms

Two-way SMS is native; email marketing is described as syncing with Klaviyo rather than sending natively. https://gotab.com/features/customer-engagement · retrieved 2026-08-01

C
Yes

Handles or explicitly documents A2P 10DLC brand and campaign registration required for the operator to send SMS through the platform.

guest-loyalty-10dlc-registration

The zero-hit on the acronym 10DLC is a vocabulary artefact: GoTab documents the substance step by step in the vocabulary the regulators themselves use. This article tells the operator to 'Complete the brand and campaign registration process through OpenPhone' and then walks thirteen numbered registration steps, including 'Enter Your Business Registration Number (EIN)', the exact use-case sentence and two sample messages to paste, a hosted consent screenshot to submit as the opt-in link, the registration fee, Trust Center approval, and multi-EIN handling for multi-location groups. The companion announcement sms-platform-transition names the scheme outright — register your brand with the Campaign Registry, register separate campaigns for each type of messaging, reviewed by an agency chosen by the major US cellular carriers — and records that GoTab carried transition traffic on its own registered brand and campaigns. So GoTab both handles it and explicitly documents it; this article also notes 'Signing up for OpenPhone is not a requirement for basic text functionality on GoTab', i.e. the GoTab-registered brand covers order-status SMS. Shortfall to carry: the registration is executed by the operator inside a separately-paid OpenPhone plan, not inside GoTab, and the guidance dates from the 2024 transition. Searched all four fetched hosts. https://success.gotab.io/knowledge/openphone-account-setup · retrieved 2026-09-05 adversarially verified

B
Partial

Reports redemption and incremental sales attributable to a specific campaign or offer, tying redeemed offers to actual check totals. differentiator

guest-loyalty-campaign-attribution

Redemption-to-check linkage exists on the Sales Page: the discounts line breaks out 'each discount and the amount discounted', filterable by discount type or reason, and 'Click download CSV and we provide a report with links to tabs for each discount applied during the selected date range' — every coupon/offer redemption ties to actual tabs. Shortfalls: no incremental-sales or lift measurement against a control group, and no send-level campaign reporting (email/SMS campaigns live in Klaviyo, whose attribution stays in Klaviyo). https://success.gotab.io/knowledge/how-to-view-your-sales-report · retrieved 2026-08-05

B
Partial

Provides self-serve export of the full guest list and transaction history, including contact PII, via CSV or API, without a vendor fee or support ticket. differentiator

guest-loyalty-data-export-portability

List customers via public API supports programmatic export; a self-serve bulk CSV of guests plus transaction history is not documented. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Partial

Triggers post-transaction feedback requests and routes responses by score, sending low scores to private service recovery and high scores toward public review sites. differentiator

guest-loyalty-review-capture-routing

Feedback capture and review exist: a dedicated 'Reports: Customer Feedback' page sits under Manager Dashboard > Reports, the Feedback page is filterable by scheduled vs placed tabs, QR ratings arrive as POS notifications, and two-way SMS plus the guest Help button give staff a recovery channel from the KDS. Shortfall: no score-based routing is documented — nothing sends low scores to a private service-recovery queue or steers high scores toward public review sites (the receipt CTA button is a single static URL, not rating-conditional). https://success.gotab.io/knowledge/how-to-view-customer-feedback · retrieved 2026-08-05

B
No

Issues an Apple Wallet or Google Wallet loyalty pass with push-updatable balance, as an alternative to requiring a branded app download. differentiator

guest-loyalty-wallet-pass

GoTab enumerates the member-facing surfaces for balance and offers and a wallet pass is not among them: 'Customers can view their membership information from their GoTab Account', reached via the user icon then Account > Settings to see available coupons and membership information, while in-venue identification is 'Rewards Check-in' by phone number at the POS or CFD. Across all four fetched hosts (877 docs) the terms apple wallet, wallet pass, passkit and pkpass return zero hits; the only two google wallet hits are FAQ answers about accepting digital-wallet payments at the kiosk, that is a tender type and not a loyalty pass. The app-alternative GoTab actually sells is RFID Pass, a physical wristband or card, a different mechanism. Falsified by any article on issuing a pkpass or a Google Wallet loyalty object, or a push-updatable balance pass in the API reference. https://success.gotab.io/knowledge/membership/loyalty-check-in-on-pos · retrieved 2026-09-05 adversarially verified

B
No

Provides admin tooling to fulfill guest data access and deletion requests (CCPA/CPRA-style) including deletion propagation to loyalty and marketing records.

guest-loyalty-privacy-rights-tooling

CLAIM: operator-facing tooling to service guest privacy rights (access / export / erasure of a guest's records). VERDICT: no, on a two-sided ENUMERATION that this cell had never been checked against - the prior note reasoned only from the KB and the Account Deletion Policy. NEW MEASUREMENT (327th): (1) GoTab's COMPLETE published REST API, https://docs.gotab.io/specs/gotab-api.json (OpenAPI 3.1, 'GoTab API' v2.0, production server https://gotab.io/api) = 65 paths, 90 operations, 72 schemas, of which 10 are DELETEs. The customer surface is exactly four operations and none of them is an erasure or an export: GET /integrations/customers ('Lists existing customers for an integration'), POST /integrations/customers ('Creates a new customer'), GET /integrations/customers/{customer_profile_uuid} ('Retrieves a customer'), and segment-membership removal only - DELETE /v2/loc/{location}/segments/{segment}/customers/{customerId} and DELETE /v2/loc/{location}/segments/{segment}/customer_handle/{customerHandle}, both described as 'removes that customer from the provided segment'. Removing someone from a marketing segment is not erasure of the guest record. The only other DELETEs are /loc/{location}/users/{user} (a STAFF user, not a guest), /wallet/payment-methods (the guest's own wallet), modifiers, schedules, tab-passes, item-groups and an integration-location link. (2) GoTab's COMPLETE published GraphQL schema, https://docs.gotab.io/graphql-schema.txt (376,674 bytes): tokenised, case-insensitive controls return 'delete' 0, 'erase' 0, 'anonymi' 0, 'export' 0, 'redact' 0, 'purge' 0, 'forget' 0, 'gdpr' 0, 'ccpa' 0, 'consent' 0 - the schema is read-shaped and contains no erasure or export field at all. (3) The only erasure route GoTab publishes remains GUEST-initiated and guest-scoped, as the standing note records (Account Deletion Policy: self-service at gotab.io/cust/account, or an email request). (4) Re-probed today with curl and a browser UA: gotab.com publishes no DPA, subprocessor list, trust page or compliance page - /dpa, /data-processing-addendum, /terms, /msa, /legal, /trust, /compliance and /subprocessors all return the SAME 66,044-byte body as the nonsense control /zzz-nonsense-ctl, and only /privacy-policy (79,954 b) and /security (69,560 b) are real pages; on gotab.io every one of those paths 302s to /not_found (3,487 b, identical to the nonsense control). So there is no contractual surface that could carry an operator DSAR obligation either. Retrieved 2026-09-18. https://docs.gotab.io/specs/gotab-api.json · retrieved 2026-09-18 adversarially verified

B
No

Vendor documents AI/ML-generated recommendations for offer content, target audience, or send timing (claim-level; verify it is a shipped product feature, not a roadmap statement). differentiator

guest-loyalty-ai-offer-recommendation

The vendor enumeration of this exact capability area has no AI in it. Under 'Features to Maximize Guest Engagement' the page lists Segments, 'Coupons and Discounts', Integrated Email Marketing and Gift Cards, and the four engagement pillars above it are Guest Segmentation, Guest Loyalty, Two-Way Text Communication and Email Marketing — manual construction throughout, with automation delegated to Klaviyo. Across GoTab-authored documents on all four fetched hosts, artificial intelligence returns zero hits and machine learning only two, both opsi.io blog posts about invoice automation, a separately-priced add-on module and not offer targeting; every AI hit on gotab.com is an /integrations/ or /partners/ page describing another company product (Pulse AI, Restoke, Marqii, TastyVox, Clutch, Tenzo), which per the subject rule cannot score a GoTab cell. Falsified by a product-update note or feature page announcing AI-generated offer copy, audience suggestion or send-time optimisation as shipped. Grade C because the instrument is vendor marketing — but the claim is itself a marketing-claim cell, so the absence of the claim from the vendor own marketing is close to dispositive. https://gotab.com/features/customer-engagement · retrieved 2026-09-05 adversarially verified

C
Yes

Offers native stored-value/gift card balances tied to the same guest profile and redeemable across all locations of the brand.

guest-loyalty-stored-value-gift

Native gift cards (+$35/mo) digital and physical, plus RFID Pass reloadable stored value redeemable at POS and kiosks. https://gotab.com/products/rfid-pass · retrieved 2026-08-01

C

Labor & workforce

Yes

Employees clock in and out at the POS terminal using a PIN, badge, or card, with no separate time-clock hardware required.

labor-clock-in-at-pos

Staff clock-ins are listed among functions that continue at the POS during an outage, confirming clock-in happens on the terminal. https://gotab.com/faq · retrieved 2026-08-01

B
Yes

Records time punches while the POS is offline and reconciles them on reconnect without loss or duplicate entries, with the behavior documented. differentiator

labor-offline-time-punch

Staff clock-ins function offline and GoTab automatically syncs transaction data, order history and inventory updates back to the cloud on reconnect. https://gotab.com/faq · retrieved 2026-08-01

B
Partial

Permissions are grantable per discrete action (void, comp, discount, refund, drawer open, price change, report access) per role and per location, not limited to fixed cashier/manager/admin tiers.

labor-granular-rbac

Public API exposes List/Get User Roles and full user CRUD including bulk update; per-discrete-action and per-location permission granularity is not documented. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Yes

Manager overrides and approvals are attributed to the individual approver and written to an immutable audit trail queryable after the fact.

labor-manager-override-audit

Manager overrides are PIN-based and the approving manager is recorded and surfaced in reporting for comps, voids and discounts — 'the manager that PIN'd in to approve'. Override plus attributable audit trail is documented. https://success.gotab.io/knowledge/service-employee-reports · retrieved 2026-08-01 adversarially verified

B
Partial

Builds and publishes employee shift schedules inside the same platform that holds the timekeeping data, rather than requiring a third-party scheduling app. differentiator

labor-native-scheduling

API has Create/Update/Archive labor schedule endpoints, but GoTab's directory positions scheduling with 7shifts, Workforce.com and Schedulefly as the primary path. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Partial

Generates recommended staffing levels or labor hours by daypart from the operator's own POS sales history, not just a static template. differentiator

labor-demand-labor-forecast

GoTab's native output stops at demand prediction: the Daily Insight email includes 'projected sales' alongside daily sales, product mix and fulfillment times, but no staffing-level or labor-hours recommendation exists in any labor or dashboard doc. The documented path to demand-driven labor is the 7shifts integration — 'When a tab (check) is closed in GoTab, the sales detail is exported into 7shifts' at line-item level with the GoTab zone as revenue_center — feeding 7shifts' scheduling product. Shortfall: forecast-to-staffing conversion happens entirely in 7shifts, a separately subscribed partner product, not in GoTab. https://success.gotab.io/knowledge/how-gotab-supports-7shifts-tip-pooling · retrieved 2026-08-05

B
Partial

Displays labor cost as a percentage of sales in real time during service on a manager view or POS, not only in end-of-day or next-morning reports. differentiator

labor-realtime-labor-percent

Manager Dashboard lets operators review cost of goods, schedule and labor costs in real time; labor as a live percentage of sales during service is not specifically documented. https://gotab.com/products/manager-dashboard · retrieved 2026-08-01

D
Partial

Warns or blocks at clock-in when an employee is approaching or would cross a configured overtime threshold, before the overtime is incurred. differentiator

labor-overtime-prevention

Delivered by the 7shifts integration, not natively - the same partner pattern this record already uses for shift swaps and labor forecasting. GoTab's own labor settings (Print/Declare Tips/Prevent Clock Out with Open Tabs/Prevent User Report with Open Tabs/Automatic Clock Out/Require Clock In, break types, labor edit reasons) contain no overtime threshold, and GoTab's only punch gate is 7shifts Clock-in Enforcement, which verifies 'the user is first clocked-in to 7Punches before allowing them to access the point of sale' - a clocked-in check with no hours logic. 7shifts supplies the overtime layer on top: Overtime Warnings and Alerts require 'a POS labor integration or 7punches enabled' plus configured daily and weekly overtime rules, and fire when employees 'enter daily overtime, weekly overtime, or are at risk of weekly overtime', with an at-risk buffer before the weekly threshold is crossed. Shortfalls: warnings are manager-facing dashboard flags and mobile push notifications, not a warning or block presented to the employee at the GoTab clock-in prompt; the punch can still be taken; the thresholds live in 7shifts, a separately contracted product. https://kb.7shifts.com/hc/en-us/articles/5130165307795-Overtime-Alerts · retrieved 2026-08-06

B
Partial

Configures meal and rest break rules per state or jurisdiction, prompts required break attestation, and flags missed breaks for premium pay. differentiator

labor-break-compliance-by-state

Delivered by the 7shifts integration, not natively - the same partner pattern this record already uses for overtime prevention, shift swaps and labor forecasting. GoTab's own labor page carries only operator-created break types with a paid/unpaid toggle and six labor policies (Print on Clock Out, Declare Tips on Clock Out, Prevent Clock Out with Open Tabs, Prevent Printing of User Report with Open Tabs, Automatic Clock Out, Require Clock In) - no jurisdiction selector, no attestation prompt, no premium-pay trigger. The break stream reaches 7shifts: its GoTab POS article states 'When an employee clocks in, their open punches are immediately sent to 7shifts. Breaks, declared tips, and punch-out times are sent to 7shifts upon punch-out', and 7shifts' Labor Exceptions report 'is only available if you use 7punches or a POS labor integration'. On top of that stream 7shifts sets Jurisdiction under Settings > Labor & Compliance (California, California - Berkeley, California - Emeryville, plus New York, Washington, Illinois, Oregon, Nevada, D.C., Philadelphia and Canada articles), computes Missed Meal Break exceptions when a required break is not taken or 'Meal Break #1 is not started before the end of the 5th hour (4:59) of the shift' and Missed Rest Break exceptions, each of which means 'an employee should be rewarded with 1 hour of pay at the regular rate of pay', and supports signed '6hrs & out' / '10hrs & out' meal-break waivers with Employee Attestation, which is mandatory for dismissing a rest break. Shortfalls: none of this is GoTab's - the rules, attestation and exception reporting live in 7shifts, an Entree-plan-or-higher product GoTab's own setup doc requires, and 7shifts adds that 'access to California labor compliance features may require an upgrade from your current 7shifts plan'; the attestation appears in the 7shifts punch modal rather than at the GoTab clock-out prompt; exceptions are reported for payroll rather than blocking or prompting the employee at the terminal; and coverage is a fixed jurisdiction list, not per-state rules for every state. https://kb.7shifts.com/hc/en-us/articles/4417505124627-California-Labor-Compliance · retrieved 2026-08-06

B
Partial

Computes tip pooling and tip-outs automatically per shift from configurable rules (percentage of sales, hours worked, points, or role), rather than requiring a manual spreadsheet.

labor-tip-pooling-rules

Handled by TipHaus and eTip integrations rather than a native rules engine. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

Retains a per-shift, per-employee record of tips received, amounts contributed to the pool, and amounts distributed from the pool, exportable for audit.

labor-tip-distribution-audit-trail

Natively GoTab records tips per employee and retains a labour audit trail, but not a POOL one: the Labor page carries 'Declare Tips on Clock Out', a Download Report of every clock in/out with pay and role, a Download Payroll Report of hours/rate/pay per user, and a Download Audit of edited punches with editor and reason - the audit covers time-punch edits, not tip allocations. Contribution and distribution live in 7shifts, and GoTab sends it more than the previous note credited: on each closed tab the whole receipt with 'tip_details', and on each punch-out the breaks, punch-out time and DECLARED tips, so all three tip sources - CC tips, auto-gratuity and declared tips - arrive. 7shifts Tip Pooling then redistributes on hours worked, points or percentages and includes the result in its reports, 7shifts Payroll and external payroll exports. Shortfalls: the pool contribution/distribution ledger and its audit export are a 7shifts artefact requiring the 7shifts Entree plan or higher plus 7punches or an Actual Labor integration, published schedules and role mapping (otherwise tips fall to an 'unassigned classification'); GoTab documents no per-shift pool audit trail of its own. https://kb.7shifts.com/hc/en-us/articles/4417504746643-GoTab-POS · retrieved 2026-08-06

B
Partial

Payroll export separates cash from charged qualified tips and carries a Treasury tipped-occupation code, supporting the W-2 Box 12 code TP and Box 14b reporting required for tax year 2026. differentiator

labor-qualified-tips-w2-reporting

The cash-versus-charged separation is documented, the W-2 coding is not, and neither happens in GoTab. 7shifts' GoTab POS article states 'The GoTab POS integration supports the following tip sources: CC Tips (Credit Card Tips), Auto-gratuity, Declared Tips' and that to include them 'in your reports and payroll (7shifts Payroll and external exports), you must configure your tip calculation settings'; declared cash tips ride the punch ('Breaks, declared tips, and punch-out times are sent to 7shifts upon punch-out'), originating in GoTab's own 'Declare Tips on Clock Out' modal, while CC tips ride the closed check. That is exactly the qualified-tip cut - voluntary tips separated from mandatory auto-gratuity, cash separated from charged. Shortfalls: none of the tax coding is GoTab's. GoTab's own Download Payroll Report carries only 'Total hours worked, rate and pay for each user' with no tip field and GoTab has no native payroll (labor-native-payroll = no); the tipped-occupation code exists only as a manually entered SOC field in the 7shifts Payroll employee profile ('Locate the field for the SOC code and enter the corresponding numeric code from the BLS website'), sourced for South Carolina and West Virginia unemployment returns rather than from GoTab, and never described as a Treasury tipped-occupation code; and the only OBBBA article ('pertains to customers using 7shifts Payroll only') describes tax year 2025 reporting where Check Technologies 'populates Box 14 of the W-2 with a Qualified OT label' and employees 'should refer to Box 7 of their W-2 to estimate their qualified tips' - no Box 12 code TP and no Box 14b emission is documented for tax year 2026, and 7shifts' own FAQ still says 'no changes are required in 7shifts for how tips or overtime are declared' under the No Tax on Tips law. The path also requires 7shifts Entree plan or higher plus the separately purchased 7shifts Payroll; GoTab's other labor partners (Workforce.com, Schedulefly, TipHaus, eTip) document no qualified-tip path. https://kb.7shifts.com/hc/en-us/articles/4417504746643-GoTab-POS · retrieved 2026-08-06

B
No

Vendor processes payroll itself as a first-party product (tax filing and direct deposit), as opposed to exporting to a third-party payroll provider only. differentiator

labor-native-payroll

No payroll product exists. GoTab's Payroll/HR/Labor category is populated entirely by third parties (7shifts, Workforce.com, Schedulefly, TipHaus, eTip, Avero) and the price sheet lists no payroll SKU. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

Exports timecards, wages, and tips in documented formats or via direct integration for at least two major payroll providers (e.g. Gusto, ADP, Paychex, QuickBooks).

labor-payroll-export-formats

GoTab exports payroll data natively but ships no payroll-provider connector. The manager dashboard Labor page offers three downloads over a date range of up to 31 days - Download Report (each user's clock ins/outs with pay and role), Download Payroll Report (total hours worked, rate and pay per user) and Download Audit (edited punches with editor and reason) - and cash tips are captured through Declare Tips on Clock Out. Shortfalls: the download file formats are not documented, so 'documented formats' is unmet; no direct integration with any payroll provider is published, and the QuickBooks Online integration is a general-ledger export (accrual only, GL code mapping), not payroll. Reaching Gusto, ADP, Paychex or QuickBooks Payroll requires the 7shifts hop - GoTab sends punches, breaks and all three tip sources to 7shifts, which runs 7shifts Payroll and external payroll exports. CORRECTED 2026-08-06: the previous note scored this off gotab.com/integration-partners, an integration directory, which is not an exhaustive enumeration and had also missed the native reports entirely. https://success.gotab.io/knowledge/howtomanagelabor · retrieved 2026-08-06

B
Partial

Employees can request shift swaps or claim open shifts in a self-service app, subject to a manager approval workflow that enforces overtime and role eligibility. differentiator

labor-shift-swap-workflow

Delivered by scheduling partners rather than natively: GoTab has no employee self-service app (its staff surface is the POS), and scheduling is documented partner territory — 7shifts is integrated to the depth of employee sync, PIN mapping, clock-in enforcement and per-check sales export (KB 7Shifts section), with Schedulefly and Workforce.com also in the directory. Shift-swap requests, open-shift claiming and manager approval therefore happen in the partner scheduler against schedules synced with GoTab labor data; GoTab documents no native swap workflow of any kind. https://gotab.com/integration-partners · retrieved 2026-08-05

B
Partial

Reports per-employee sales performance metrics (average check, items or category attachment rate, void and comp rate) by server or cashier. differentiator

labor-server-performance-metrics

POS page says operators can monitor employee performance; specific per-server metrics are not enumerated. https://gotab.com/products/point-of-sale-pos · retrieved 2026-08-01

D

Inventory, purchasing & cost control

Partial

Supports multi-level recipe/BOM costing where a recipe can contain sub-recipes (prep items) and plate cost recalculates automatically when any component ingredient's cost changes.

inventory-recipe-bom-costing

Delivered by partners — opsi recipe/inventory modules at +$45/mo each on GoTab's price sheet, plus meez, Craftable, MarginEdge, WISK. Multi-level sub-recipes are not a documented GoTab-native capability. https://gotab.com/pricing · retrieved 2026-08-01

C
Partial

Stores purchase unit, recipe unit, and count unit per item with explicit conversion factors, plus a yield/waste percentage applied to raw-to-usable conversion.

inventory-unit-conversion-yields

Core GoTab inventory is count-based only — stock, par, reserved and available with recount/received/loss/theft adjustments and no unit concept at all (success.gotab.io/knowledge/inventory-management). Unit handling lives in opsi, the recipe/food-costing module GoTab acquired in March 2025 and prices as a separate $45/mo add-on, whose feature page states 'with built-in unit conversions and yield percentages, you can skip the math headaches.' Shortfalls: requires the separately priced opsi add-on rather than the core POS, and the feature page does not document the purchase/recipe/count unit triple or how conversion factors are configured. https://www.opsi.io/products/food-costing · retrieved 2026-08-04

C
Partial

Produces a theoretical-vs-actual usage variance report per item for a defined count period, showing theoretical usage from POS sales times recipe, actual usage from counts plus receipts, and the variance in both units and currency. differentiator

inventory-theoretical-vs-actual

Variance reporting exists only in opsi, the GoTab-owned $45/mo add-on: its inventory product advertises 'Export category level variance reports... analyze trends and spot discrepancies before they snowball into lost profits', with counts of both products and recipes and invoice-synced costs. Shortfalls: category-level, not the per-item theoretical-vs-actual variance in units and currency the claim specifies; the theoretical side (POS sales times recipe) is not spelled out on the page; and it requires the opsi module — core GoTab inventory records only counts and adjustments (recount/received/loss/theft) with no recipe linkage. https://www.opsi.io/products/inventory · retrieved 2026-08-05

C
Partial

Depletes ingredient on-hand quantities in near real time as orders are fired at the POS, including depletion driven by modifier selections, not only in an end-of-day batch. differentiator

inventory-realtime-depletion

API exposes product-variant inventory, SKU stock level updates and Apply inventory instructions; modifier-driven depletion and near-real-time fire-time decrementing are not documented. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Partial

Automatically 86s (marks unavailable) menu items on the POS, online ordering, and connected third-party delivery menus when a component ingredient hits zero or a configured threshold. differentiator

inventory-86-auto-sync

Auto-86 fires from product stock, not from ingredients: a tracked product's out-of-stock action ('86 Product: When stock levels hit 0, the product will be temporarily disabled and automatically re-enabled the following day') marks it out of stock on POS and the customer-facing ordering menu, and the 86 state resyncs to DSP menus through the 3rd Party Ordering aggregator (its FAQ attributes rejected orders to 'resyncing is not yet complete across all platforms'). Shortfalls: the trigger is product/variant stock hitting zero — no component-ingredient or recipe-level threshold exists in core GoTab (recipes live in the opsi add-on) — and 3P propagation is laggy with no documented latency. https://success.gotab.io/knowledge/inventory-management · retrieved 2026-08-05

B
Partial

Supports full physical inventory, ad-hoc spot counts of a subset, and scheduled recurring cycle counts, with each count's variance recorded as separate history rather than overwriting prior counts.

inventory-count-modes

POS supports inputting recounts and recording received items; full vs spot vs scheduled cycle counts with separate variance history are not documented. https://gotab.com/products/point-of-sale-pos · retrieved 2026-08-01

D
Partial

Provides a mobile/tablet counting app supporting barcode or QR scanning that continues to accept counts with no network connectivity and syncs on reconnect.

inventory-mobile-count-offline

FAQ confirms inventory tracking continues offline and syncs on reconnect; a barcode/QR-scanning count app is not documented. https://gotab.com/faq · retrieved 2026-08-01

B
Partial

Ships pre-built electronic vendor catalog and order-transmission integrations with named broadline distributors (e.g. Sysco, US Foods, Performance Food Group) that send purchase orders and receive invoices electronically rather than by emailed PDF. differentiator

inventory-vendor-catalogs-edi

Buyers Edge Platform and Buyers Edge Back Office are named partners, but no named broadline EDI link to Sysco, US Foods or PFG is documented. https://gotab.com/integration-partners · retrieved 2026-08-01

E
Partial

Ingests supplier invoices from photo, PDF, or email and extracts line-item detail (item, quantity, unit, price) into the inventory ledger without manual line entry. differentiator

inventory-invoice-ocr

Available through MarginEdge, Restaurant365, opsi and Craftable partners; not a native GoTab capability. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

Tracks per-item purchase price history across invoices and alerts when a received price exceeds the prior or contracted price by a configurable threshold. differentiator

inventory-price-change-alerts

Exists in the opsi add-on, not core GoTab: opsi's food-costing page documents 'detailed ingredient cost history tracking', vendor-invoice syncing for 'real-time pricing updates', and threshold alerts — 'set custom price thresholds and receive notifications the moment ingredient costs fluctuate'. Shortfalls: requires the separately priced opsi module (+$45/mo on GoTab's price sheet); comparison against a contracted price is not documented; and the evidence is a feature page rather than an admin guide. Core GoTab inventory tracks counts only, with no purchase-price concept at all. https://www.opsi.io/products/food-costing · retrieved 2026-08-05

C
Partial

Maintains par levels per item per location and generates suggested purchase order quantities from on-hand versus par, with at least one mode driven by a sales forecast rather than a static par. differentiator

inventory-par-auto-suggest

Par levels exist per product but do nothing beyond flagging: 'Par Level: The inventory needed to fulfill demand. This will make the stock level red if the stock count falls below the par level.' Shortfalls: no suggested purchase-order quantities are generated from on-hand vs par, and no forecast-driven mode exists — neither core GoTab nor the opsi module lineup (recipes, food costing, invoices, bill pay, task lists, counts, nutrition labels) documents any purchasing or ordering feature. https://success.gotab.io/knowledge/inventory-management · retrieved 2026-08-05

B
Partial

Provides a structured waste/spoilage logging workflow with reason codes that debits inventory and reports waste cost separately from usage variance.

inventory-waste-logging

POS page describes accounting for loss and theft; structured reason codes and separate waste-cost reporting are not documented. https://gotab.com/products/point-of-sale-pos · retrieved 2026-08-01

D
No

Tracks expiration or use-by dates on received and prepped items and surfaces an expiring-soon report or alert before spoilage.

inventory-shelf-life-expiry

Positive evidence of absence, from GoTab's own generated data model (retrieved 2026-09-15; the dump stamps itself 'Generated: 2026-09-14T18:05:07.151Z'). The published GraphQL schema parses to 349 type blocks, and the inventory entities are closed, machine-generated column lists carrying no date but the record's own lifecycle timestamps: `Inventory` (39 fields) = inventoryId, name, locationId, created, archived, sku, type, units, impactsAvailability, detached, tracked, inventoryLevel, stockLevel, parLevel, reservedQuantity, idealReservedQuantity, lowLevelComponents, nextLevelComponents, instructionComponents, pathComponents, parentProductId, variant, conversionFactor, xSkuIds, externalInventoryId, minusOneComponents, plusOneComponents, kdsVisible, routerIds, isPrepRequired (+relations); `VariantInventory` (28) adds only instructionId, price, externalSkuIds, externalSku, barcodes; `InventoryAdjustment` (12) = inventoryId, created, transactionType, orderId, adjustment, confirmedQuantity, inventoryAdjustmentId, adjustmentReason, cost (+2 relations). The strongest census artefact is the generated sort/filter surface, which enumerates every underlying COLUMN rather than merely the exposed fields: `InventoriesOrderBy` runs NATURAL, INVENTORY_ID, NAME, LOCATION_ID, CREATED, ARCHIVED, SKU, TYPE, UNITS, IMPACTS_AVAILABILITY, DETACHED, TRACKED, INVENTORY_LEVEL, STOCK_LEVEL, PAR_LEVEL, RESERVED_QUANTITY, IDEAL_RESERVED_QUANTITY, the five *_COMPONENTS, PARENT_PRODUCT_ID, VARIANT, CONVERSION_FACTOR, X_SKU_IDS, EXTERNAL_INVENTORY_ID, KDS_VISIBLE, ROUTER_IDS, IS_PREP_REQUIRED, PRIMARY_KEY (each ASC/DESC) - CREATED and ARCHIVED are the only date columns, and `InventoryCondition` / `VariantInventoryCondition` / `InventoryAdjustmentCondition` list the identical sets. The adjustment vocabulary is a closed enumeration with no date-driven or spoilage member: `InventoryAdjustmentReasonType` = [RECOUNT|RECEIVED|LOSS|THEFT|TRANSFER|ORDER|SCHEDULED_ORDER|VOID|PREP] and `InventoryAdjustmentType` = [ORDER|SCHEDULED_ORDER|VOID|INVENTORY_RECEIPT|INVENTORY_COUNT|ADJUSTMENT|HELD_ORDER]; `InventoryUnits`/`ItemInventoryUnits` are 16 quantity measures (COUNT, OZ, LB, KG, G, FLOZ_, LITRE, ML, GAL, QT, PINT, CASE, DRY_PT_, CANS, BOTTLES, SMIDGE) with no date dimension, and `ImpactsAvailabilityType` = [EIGHTYSIX|DISABLE]. This is decisive for the claim's exact wording, because RECEIVED and PREP adjustments are the very 'received and prepped items' it asks about, and the record written for each is a 12-field row whose only timestamp is `created`. The REST spec agrees: OpenAPI 3.1 'GoTab API' 2.0, 65 paths / 90 operations, whose inventory endpoints (PATCH /loc/{location}/inventory/skus/{sku}, POST /loc/{location}/inventory/recipes, POST /loc/{location}/inventory/{sku}/transfer, GET|PATCH /loc/{location}/inventory/{variant}, GET|POST /loc/{location}/products/{product}/inventory) carry adjustment_reason [RECOUNT|RECEIVED|LOST|THEFT] and [RECOUNT|RECEIVED|LOSS|THEFT|TRANSFER], and the only temporal property anywhere under an inventory schema is `updatedAt` on InventoryItem/InventoryProduct - a row-modification stamp, not a food date. The manager dashboard matches the model rather than exceeding it: GoTab's own screen documentation for the exact inventory surfaces describes Manage Stock as Recount / Received / Loss / Theft, Inventory Details as Par Level / Out-of-Stock Action / Untrack, and the depletion report (docs.gotab.io/operator/manager-dashboard/reports-inventory-depletion/) as a date-range-and-adjustment-type filter returning Sku, Product, Total Adjustments, Units - no use-by field on receiving and no expiring-soon view. CONTROLS/POISON: swept TOKENISED, not by substring. Whole-token counts across the schema are {expires: 1}, that one being a comment on `Product.manualDelayUntil` ('Timestamp when manual delay expires and auto-calculation should resume') - a prep-time override, not food dating; in the REST spec {expire: 5, expires: 6}, all of them `expire_month`/`expire_year` (CARD expiry on /wallet/payment-methods) and `expires`/`expiresIn` (OAuth token, 86400s). A naive substring match on 'expir' would have manufactured a false yes from all three. shelf, spoil, spoilage, perishable, fifo, lot, batch, waste and freshness return zero whole-token hits in both specs. The same tokenised sweep over the 334-page docs corpus returns only OAuth/payment/loyalty expiry (`expiration_date` on loyalty points) plus one 'rotation' that is automatic SCREEN rotation in a Guest Pay release note, and over the 170-file served bundle only card/token expiry and CSS rotation. BOUND, stated plainly: this is core GoTab's published data model plus its own dashboard screen docs. It does not cover opsi, the GoTab-owned but separately priced (+$45/mo) recipe and inventory module where this record's recipe, costing and variance capability lives; opsi's feature pages take up dates in no form, so there is no positive evidence of shelf-life tracking there either, but a future opsi data-model source could move this cell to partial rather than no. https://docs.gotab.io/graphql-schema.txt · retrieved 2026-09-15 adversarially verified

A
Partial

Supports partial-bottle liquid inventory by weight (scale integration) or by bottle fraction for bar and beverage counting.

inventory-bar-partial-bottle

Beverage-specific partners BarTrack, Draft Control, WISK and Smart Bar USA cover keg/draft flow and liquid counting; native partial-bottle-by-weight is not a GoTab feature. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

Exports period COGS and accounts-payable invoice detail in a named accounting system's import format (e.g. QuickBooks, Sage Intacct, NetSuite) with GL account mapping configurable per item category.

inventory-cogs-gl-export

QuickBooks Online sync (native; claimed at no extra cost in the Toast comparison though the price sheet shows accounting integration at +$10/mo on Basic), plus Shogo, Restaurant365 and DAVO. Per-category GL account mapping is not documented. https://gotab.com/integration-partners · retrieved 2026-08-01

B
No

Inventory and recipe costing are delivered natively by the POS vendor rather than requiring a separately-contracted third-party product (e.g. MarginEdge, MarketMan, Restaurant365) to function. differentiator

inventory-native-not-partner

GoTab has basic native stock tracking, but recipe costing and full inventory management are separately contracted — the price sheet itemizes opsi invoice/recipe/inventory/task modules at +$45/mo each. https://gotab.com/pricing · retrieved 2026-08-01

C
Partial

Joins recipe cost to actual item sales mix to report contribution margin per menu item, and flags items whose margin fell below a configured threshold after ingredient cost changes. differentiator

inventory-menu-margin-linkage

Margin analysis lives in the opsi add-on: 'Cost summary reports... highlight menu prices, profit margins, and overall food cost percentages' and 'Historical profitability analysis... visualizes recipe profitability trends so you can make data-backed decisions', with invoice-synced ingredient costs keeping calculations current. Shortfalls: requires the $45/mo opsi module; the documented alerts fire on ingredient price thresholds, not on an item's margin falling below a configured threshold; and the join to actual POS sales mix (versus recipe-level costing) is not explicit on the page. https://www.opsi.io/products/food-costing · retrieved 2026-08-05

C

Reporting, BI & data access

Yes

Provides a live sales dashboard reflecting transactions within minutes of the sale, accessible from both a browser and a mobile app off-premise.

reporting-realtime-dashboard

Manager Dashboard gives real-time and historical sales, accessible from computer or mobile off-premise, across multiple venues. https://gotab.com/products/manager-dashboard · retrieved 2026-08-01

C
Partial

Generates a daily close-out/EOD report reconciling gross sales, net sales, tax, tips, discounts, refunds, tender types, and expected cash deposit in one document.

reporting-eod-closeout

The nightly pieces exist but no single closeout document: the Sales Page drills into every line (discounts by type/reason with tab-linked CSVs) and the Transaction Report CSV carries payment type, tip and auto-gratuity per payment; cash accounts reconcile counted-vs-expected to the Balance Accounts page with over/short flowing into reporting; the Payouts dashboard shows per-fiscal-day deposits; a Daily Insight email snapshots sales, product mix, projected sales and fulfillment times; and the KDS prints its own closeout of items, subtotals and tips. Shortfall: no documented single EOD packet reconciles gross/net sales, tax, tips, discounts, refunds, tenders and expected cash deposit in one artifact. https://success.gotab.io/knowledge/how-to-view-your-sales-report · retrieved 2026-08-05

B
Partial

Provides a product mix (PMIX) report at individual menu item AND modifier level with quantity, gross and net sales, filterable by daypart and revenue center.

reporting-pmix-modifier-level

The Product Mix report shows gross and net quantity and gross and net sales (voids excluded from both, comps in gross netting to zero), with a configurable drilldown hierarchy that includes options — GoTab's term for modifiers — as a drill level ('fiscal day breakdown, then accounts, products, and options'; 'Options must be the last drilldown option'). Shortfall: documented PMIX filters are date range, placed-versus-scheduled and user only — no daypart or revenue-center/zone filter is documented on the report. https://success.gotab.io/knowledge/how-to-view-your-product-mix · retrieved 2026-08-04

B
Yes

Produces an audit report of comps, voids, discounts, and price overrides attributed to the employee who applied them and the manager who approved them, with reason codes and timestamps.

reporting-comps-voids-audit

Service (Employee Reports) breaks out Comps, Voids and Discounts separately, each with initiating employee, item count, percent of sales, sales value, and the approving manager's PIN identity. That is a comps/voids audit report. https://success.gotab.io/knowledge/service-employee-reports · retrieved 2026-08-01 adversarially verified

B
Yes

Reports cash over/short per drawer, per shift, and per employee, comparing counted cash to expected cash from tendered transactions and paid-outs.

reporting-cash-over-short

Documented drawer reconciliation with variance capture: closing a cash account means 'entering the amount counted versus the amount expected and we then record that on the Balance Accounts page'; 'Any over/under will be included in your cash reporting as paid ins or outs. You can find these on your POS user reports, cash accounting and on the sales page.' Accounts are per-employee ('multiple servers/bartenders are able to operate having their own cash account to their name... they are only responsible for the cash they receive') and per-shift via the Require Reconciliation toggle (the doc's distinct AM/PM shift-accounts example), giving over/short by drawer, shift and employee. https://success.gotab.io/knowledge/how-to-reconcile-your-cash-accounts · retrieved 2026-08-05

B
Partial

Reports sales per labor hour and labor cost as a percentage of sales by hour, by department, and by employee, using actual clocked hours from the POS time clock.

reporting-labor-productivity

Dashboard surfaces schedule and labor costs alongside sales; sales-per-labor-hour by hour/department/employee is not documented. https://gotab.com/products/manager-dashboard · retrieved 2026-08-01

D
Yes

Reports per-server metrics including average check, items per check, attachment rate for named categories, and tips as a percentage of sales. differentiator

reporting-server-scorecards

The dossier scored this partial-claimed off a vague 'monitor employee performance' line on the POS marketing page. The actual metric set is enumerated in product docs: per-employee sales with item quantity, guest count, percent of sales and total sales, plus comp/void/discount counts and percentages by employee, with server reports also listed under 'Server Necessities'. https://success.gotab.io/knowledge/service-employee-reports · retrieved 2026-08-01 adversarially verified

B
Partial

Reports revenue and margin by fulfillment channel — dine-in, direct online, first-party delivery, and each third-party marketplace separately — net of marketplace commission fees. differentiator

reporting-channel-profitability

Channel revenue attribution exists — zones are the channel dimension (dine-in, takeout, delivery, and a dedicated 3rd-party zone), the Transaction Report CSV carries Zone on every payment, and the 7shifts export maps the GoTab zone to revenue_center — but profitability by channel does not: no per-zone margin report is documented, and marketplace commissions can never be netted because 'payments and payouts for 3rd party ordering are handled 100% independently of GoTab'. Shortfall: raw zone-tagged transactions only; no channel P&L and no commission or fee data in-platform. https://success.gotab.io/knowledge/how-to-view-your-sales-report · retrieved 2026-08-05

B
Partial

Supports scheduling any report for automatic recurring delivery by email or to a shared destination on a defined cadence and recipient list.

reporting-scheduled-delivery

One scheduled report exists: the Daily Insight email 'provides a snapshot of your daily sales, product mix, projected sales, and fulfillment times', sent daily to an operator-managed contact list (Location Configurations > Daily Emails), with customizable content and on-demand resend for missed dates. Shortfalls: fixed daily cadence and a single fixed report — there is no facility to schedule an arbitrary report (PMIX, labor, fulfillment, payments) on a chosen cadence to a chosen recipient list. https://success.gotab.io/knowledge/daily-emails · retrieved 2026-08-05

B
Yes

Publishes a documented public REST or GraphQL API covering orders, payments, menu, and labor data, with self-service developer credentials obtainable without a signed partner agreement. differentiator

reporting-public-api

Upheld on the score — gotab.readme.io is genuinely public with no login, and the endpoint inventory the dossier cites is accurate: I confirmed 'Adjust or transfer inventory', 'Create/Update/Archive labor schedule', 'Bulk create products', 'Bulk update products', the full Wallet set, 'Trigger OTP Challenge'/'Verify OTP Code' and 'Refund a tab payment'. The note is wrong on one point: credential creation is not self-serve (see extensibility-free-sandbox). https://gotab.readme.io/llms.txt · retrieved 2026-08-01 adversarially verified

B
Partial

Emits outbound webhooks for order and payment lifecycle events to a customer-supplied endpoint, with documented retry behavior and payload signature verification. differentiator

reporting-webhooks

GoTab does push to a customer-supplied endpoint, but not under the word webhook: this page describes an 'events-driven API to be consumed by a third party loyalty program' whose stated dependency is a 'URL for GoTab to post loyalty events to (e.g. myloyaltyplatform.com/gotab/events)', and the ACCRUAL event is 'an asynchronous request that will be triggered for each tab at a location as they are closed and updated in GoTab', sent 'for every closed tab' with full tab_data including orders, items and payments. Order and payment lifecycle data is therefore genuinely pushed. The other two conjuncts fail: authentication is a static shared secret, 'Authorization header value for your platform to confirm requests are coming from GoTab', not a payload signature, and the word retry does not occur anywhere in this document (corpus-wide the token retry hits only three readme pages, all about OAuth token refresh and accounting fiscal-day polling; Retry-After and idempotenc return zero across all 877 docs). Scope is partial too, since the push exists only for a configured loyalty or promo integration, not as an operator-subscribable event feed. Falsified by any readme page documenting retry/backoff or an HMAC signature header. https://gotab.readme.io/reference/gotab-loyalty-api.md · retrieved 2026-09-05 adversarially verified

B
Partial

API and raw-data access are included in the standard subscription rather than gated behind an additional per-location fee, revenue share, or enterprise-only contract. differentiator

reporting-api-not-upcharged

The published price sheet lists no API or data-access line item and the vendor markets an Open API, but no page affirmatively states API access is included at no extra fee. https://gotab.com/pricing · retrieved 2026-08-01

D
Partial

Core reporting beyond basic sales summaries — labor vs sales, PMIX, comps/voids audit, multi-location comparison — is available on the entry-level paid plan rather than requiring a higher tier or an add-on analytics SKU. differentiator

reporting-tier-paywall

Basic at $15/mo excludes accounting integration (+$10), multi-location management (+$50), loyalty (+$35) and 24/7 support (+$35). Whether PMIX, comps/voids audit and labor-vs-sales sit inside Basic is not documented. https://gotab.com/pricing · retrieved 2026-08-01

C
No

Documents a historical retention window of at least 24 months of transaction-level detail queryable in the reporting UI, with no silent truncation or archive retrieval fee. differentiator

reporting-history-retention

This is a documentation no, not evidence of truncation. The KB article defining the reporting surface says 'The Sales dashboard allows you to analyze real time sales data and download reports' and refers throughout to a user-chosen window, 'Click download CSV to view a report with links to tabs for each discount applied during the selected date range', but states no bound on how far back that range may go. Searching all five surfaces for retention|retain|archiv|how far back|historical data|years of data|date range yields no first-party statement of a reporting retention window: the retention hits on gotab.com are site-wide privacy and terms boilerplate plus marketing about guest retention, and the readme API docs bound queries by fiscal day and offset, never by age. No archive-retrieval fee is documented either, because no retention policy is documented at all. Falsified by any KB, pricing or merchant-terms clause naming a months or years limit or an archive fee. https://success.gotab.io/knowledge/getting-started-sales · retrieved 2026-09-05 adversarially verified

B
No

Sends proactive push or email alerts when an operator-configured metric breaches a threshold or deviates from its historical pattern (e.g. void spike, sales below forecast, discount rate anomaly). differentiator

reporting-anomaly-alerts

330th: MOVED to `no`. NEW SURFACE that defeats the standing 'the generators are server-side data behind login' objection: the GoTab Manager App KB article ships a SCREENSHOT OF THE NOTIFICATION SETTINGS SCREEN ITSELF ('Manager App Notifications Options (2).jpg', downloaded at 1,128 px and read as an image). The screen enumerates the notification types a manager may switch on, each with an Off/Mobile toggle and NO numeric or comparison input: 'Product 86'd - Get notified when a product has been 86'd'; 'POS refund requests - Get notified when a comp, void, or refund is requested'; 'Tab ratings - Get notified when a rating is received'; 'QR Scanned - Get notified when a QR code is scanned'. The capture runs to the page footer ('(c) Copyright 2024 GoTab, Inc.'), so the list is seen to its end. Every entry is a per-EVENT push; none is an operator-configured metric threshold and none compares a metric to a historical pattern, which is what this claim asks for. Corroboration, all re-measured: the article's prose adds only 'In app and push notifications. Turn on Manager App Notifications - Click the bell icon. Choose notifications you'd like to receive.'; the complete 304-page KB (href fixpoint, 825,520 chars) has `anomal` 0, `variance` 0, `low stock` 0, and its 5 `threshold` hits are a gift-card PIN threshold (3) and OpenTable's own spend-threshold alerts (2, a partner's feature inside the OpenTable app); the complete published OpenAPI (65 paths/90 operations) has no notification, alert, subscription or threshold endpoint; the 1,653-page gotab.com census has `anomal` 1, belonging to the Tenzo BI partner. I re-fetched the app shell today (manager.gotab.io, 16,773 b, 59 JS assets from cdn.gotab.io/static, 3.5 MB downloaded) and confirm the client still only does GET/PATCH /api/v2/loc/{loc}/users/{user}/configs/notifications and destructures {settings, generators} - i.e. the list is server-supplied and the screenshot is the only published enumeration of it. HONEST WEAKNESS: the screenshot's top edge is cropped just above 'Product 86'd', so a type scrolled above it cannot be excluded; the verdict rests on the fact that every visible control is a binary Off/Mobile toggle with no threshold entry anywhere in the screen. Retrieved 2026-09-18. https://success.gotab.io/knowledge/gotab-manager-app · retrieved 2026-09-18 adversarially verified

B
No

Offers a natural-language query or AI assistant that answers ad-hoc questions against the operator's own sales and labor data and returns a figure or chart, not a generic help-article response.

reporting-nl-query

GoTab positions analytics intelligence as something partners supply, not something the Manager Dashboard does: this page is headed 'Our Hospitality Business Intelligence Integration Partners' and describes Tenzo, Five x 5, Solink, Pulse AI and ProfitRover rather than a GoTab feature. Under LIMIT 2 those are other companies products and cannot score this cell. Across all five surfaces the regex natural language|AI assistant|copilot|chatbot|conversational returns exactly two documents, /media and /guides/tech-powered-hospitality, both editorial; the 21 documents containing the token AI are integration and partner pages (pulse-ai, restoke-ai, tastyvox, tenzo, marqii, wisk, clutch) plus an opsi press release. No KB article and no readme page documents an ask-a-question surface over tenant data. Grade C because the citation is a marketing page; the absence itself is measured across the closed 292-article KB and the 102 readme pages. Falsified by any KB article or product page showing a GoTab-native query box that returns a figure or chart. https://gotab.com/partners/business-intelligence · retrieved 2026-09-05 adversarially verified

C
No

Reports guest-level analytics including new versus returning guest counts, visit frequency, and lifetime spend cohorts tied to identifiable guest records from loyalty or online ordering. differentiator

reporting-guest-cohorts

Re-homed onto the enumerating document, and the value survives. The Manager Dashboard category page lists the dashboard's entire Reports set - "Service (Employee Reports)", "Reports: Sales Page", "Reports: Product Mix", "Reports: Customer Feedback", "Reports: KDS Fulfillment Report", "Reports: Payments Page", "Chart of Accounts", "Reports: Variants" and "Reports: Inventory Depletion" - nine reports with no guest report among them and no guest dimension anywhere in the dashboard's own index. The guest-side constructs are discount plumbing rather than analytics: success.gotab.io/knowledge/segments defines a segment as a group of people that discounts are associated with, enrolled manually or by phone-number upload. The Sales-page transaction export carries spot, zone and server columns but no customer id, visit count or lifetime value. Searched all four fetched hosts for returning guest, repeat guest, lifetime spend, visit frequency and cohort: six documents, all gotab.com marketing and guides plus the OpenTable page, which is a third party's guest history. Falsified by a new-versus-returning or visit-frequency report in the Manager Dashboard. https://success.gotab.io/knowledge/manager-dashboard · retrieved 2026-09-05 adversarially verified

B
Partial

Produces a forward sales forecast at daypart or hourly granularity from the location's own history, exposed in the reporting UI and consumable by scheduling or ordering workflows. differentiator

reporting-sales-forecast

GoTab natively produces one forward figure and nothing more: this KB article says GoTab offers a 'daily insight email that provides a snapshot of your daily sales, product mix, projected sales, and fulfillment times', and the token projected occurs in only two documents in the entire 877-document corpus (this one and a business-plan template guide). That satisfies a forward sales figure derived from the location history, but fails every other conjunct: it is an email digest, not the reporting UI; no daypart or hourly granularity is documented; nothing shows it feeding scheduling or par ordering; and the article never says how the projection is computed. Real forecasting is sold as a partner capability, since gotab.com/partners/business-intelligence credits ProfitRover with intelligent forecasting, which LIMIT 2 bars from scoring this cell. Falsified upward by a KB article showing an hourly or daypart forecast in the dashboard; falsified downward if projected sales proves to be a pace-to-date figure rather than a forecast. https://success.gotab.io/knowledge/daily-emails · retrieved 2026-09-05 adversarially verified

B
Partial

Generates payroll- and tax-ready tip reporting: declared versus charged tips per employee, tip pool distribution detail, and a tax liability summary by jurisdiction.

reporting-tip-tax-compliance

Documented: 'Declare Tips on Clock Out' prompts servers to enter cash tips at clock-out, a downloadable Payroll Report totals hours, rate and pay per user, and charged tips plus auto-gratuity appear per transaction in the sales reports' Transaction Report CSV. Shortfalls: no per-employee declared-versus-charged tip comparison is documented, tip-pool distribution detail is documented only through the 7shifts integration ('How GoTab supports 7shifts Tip Pooling'), and no tax-liability summary by jurisdiction exists in the documented report set. https://success.gotab.io/knowledge/howtomanagelabor · retrieved 2026-08-04

B

Hardware & physical footprint

Yes

The POS software runs on commodity off-the-shelf hardware (standard iPads, Android tablets, or PCs) that the operator can buy from a third party, rather than only on vendor-supplied proprietary terminals. differentiator

hardware-commodity-devices

No, hardware is not required. We try to use all of your existing receipt printers, tablets, and KDS systems — runs on commodity Android/iOS tablets and phones and on Windows PCs. https://gotab.com/faq · retrieved 2026-08-01

B
Partial

Vendor documentation names the specific client operating systems supported (iOS/iPadOS, Android, Windows, Linux/web) and states minimum OS versions and device specs.

hardware-os-platforms

Windows, iOS and Android are named on the pricing and FAQ pages; minimum OS versions and device specs are not published. https://gotab.com/pricing · retrieved 2026-08-01

C
Partial

Offers a purpose-built handheld ordering device with an integrated card reader (not a phone case add-on), with published drop rating and IP ingress rating.

hardware-handheld-purpose-built

The handheld is a Samsung A17 consumer phone purpose-built for GoTab's POS app — the card reader is not integrated (pair it with a separate terminal for chip and swipe), and no drop or IP rating is published. https://gotab.com/products/pocket-phone-only-pos · retrieved 2026-08-01

C
No

The handheld supports hot-swappable or field-replaceable batteries, or the vendor publishes a rated battery life for a full shift. differentiator

hardware-handheld-battery-swap

Neither disjunct is published. The handheld product page’s only shift-endurance answer is to "add a protective charging case for busy shifts" on a Samsung A17 consumer phone; the gotab.com homepage offers only an unquantified long-battery-life phrase, with no figure. The View Fact Sheet CTA that the prior verifier could not resolve is now resolved and it does not help: it is not a link but a div opening a HubSpot lead-gen popup headed "Download The Fact Sheet", so any rated figure it carries is form-gated, not published. Surfaces searched: all of gotab.com (422), success.gotab.io (275), gotab.readme.io (102) and opsi.io (59) via q210-grep - 'hot.swap|swappable' returns 0 documents and 'battery' returns 4, none of them a rating. Overturned by a public spec sheet or KB article giving rated hours, or by any statement that the battery is user-replaceable. https://gotab.com/products/pocket-phone-only-pos · retrieved 2026-09-05 adversarially verified

C
Partial

Handhelds support cellular/LTE connectivity as a documented fallback when store Wi-Fi drops, either built into the device or via a vendor-supplied failover.

hardware-handheld-lte

Cellular on the handheld is documented, but not as a Wi-Fi-drop fallback: the page offers to "use an optional SIM card for off-site events", which is vendor-supplied cellular positioned for off-premise catering, with no failover behaviour, no automatic switchover and no carrier or plan detail. Corpus-wide '\bLTE\b|cellular|SIM card|\b4G\b|\b5G\b' returns 8 documents; none describes cellular as an outage path for the venue, and GoTab's documented outage answer is the OPServer offline mode. This is deliberately consistent with the standing reliability-cellular-backup row (partial, same URL) and does not overwrite it. Overturned by any document describing automatic LTE failover on Wi-Fi loss. https://gotab.com/products/pocket-phone-only-pos · retrieved 2026-09-05 adversarially verified

C
Yes

Terminals continue to take orders, print to kitchen, and record cash sales with the internet down, and the vendor documents explicitly which functions degrade (card auth, loyalty lookup, refunds).

hardware-offline-mode

OPServer keeps terminals ordering, printing and taking cash and terminal card payments offline, with QR ordering, third-party labor and 3P delivery documented as unavailable. https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-08-01

C
Yes

Sells or supports a dedicated kitchen display system with bump-bar or touch input, station routing, and course/fire timing, as a first-party product rather than an integration-only dependency.

hardware-kds

First-party KDS sold at +$25-$35/mo with bump bar, smart prep times, colour-coded timers and zone routing. https://gotab.com/products/kitchen-display-system-kds · retrieved 2026-08-01

C
Partial

Works with standard ESC/POS thermal receipt and kitchen printers from more than one manufacturer (e.g. Epson and Star), including LAN/Ethernet printers, rather than only vendor-branded printers.

hardware-printer-compatibility

GoTab states it works with existing receipt printers, and Print Services is a tracked status-page component; no published multi-manufacturer compatibility list (Epson/Star, LAN models). https://gotab.com/faq · retrieved 2026-08-01

B
Partial

Supports the standard peripheral set — cash drawers (including multi-drawer per terminal), barcode scanners, scales for by-weight items, and customer-facing displays — with a published compatibility list.

hardware-peripherals

Re-homed off the FAQ phrase the previous note mis-attributed, onto GoTab's own public Shopify product registry, and the value survives. The registry prices three of the four categories as first-party SKUs: "Cash Drawer" in 13-inch and 16-inch variants, which for auto-kick "connnects via RJ11 cable from the cash drawer to the back of a printer" and the entry notes that auto-kick requires a printer; "Zebra DS2208 Barcode Scanner", which "connects via USB to main POS display"; and customer-facing display hardware, "ProMetal Tablet Display" plus the Swan 2 Pro CFD bundles, with the KB documenting the CFD as a software display type on any GoTab device (success.gotab.io/knowledge/displaysetup). Held to partial for three reasons: there is no scale SKU anywhere in the 78-product registry, and by-weight pricing works only indirectly through label-printing scales that emit a price-embedded barcode (success.gotab.io/knowledge/price-embedded-barcodes); nothing addresses multi-drawer per terminal, the documented pattern being one drawer behind one printer; and a catalogue of what GoTab sells is not the published third-party compatibility list the cell asks for. Falsified by a peripheral compatibility matrix or a multi-drawer article. https://shop.gotab.io/products.json?limit=250 · retrieved 2026-09-05 adversarially verified

A
Partial

Card entry happens on a PCI-listed PTS device using validated point-to-point encryption, so the POS application and tablet stay out of PCI scope; the vendor states the merchant's SAQ type.

hardware-p2pe-terminal

Card entry occurs on Adyen payment terminals (PCI PTS devices) and GoTab stores only an encrypted token, never the PAN; no validated P2PE listing and no stated merchant SAQ type. https://gotab.com/gotab-limited-warranty · retrieved 2026-08-01

B
Partial

Supports contactless acceptance directly on a commodity phone or tablet (Tap to Pay on iPhone / Tap to Pay on Android) with no separate reader. differentiator

hardware-tap-to-phone

Contactless digital-wallet payments are initiated from the handheld phone; GoTab does not name Tap to Pay on iPhone or Android tap-to-pay as the underlying capability. https://gotab.com/products/pocket-phone-only-pos · retrieved 2026-08-01

D
Yes

Hardware prices are published per SKU on the vendor's public site, rather than quote-only. differentiator

hardware-pricing-transparency

Overturned to yes. shop.gotab.io is GoTab's own public Shopify storefront, linked as Shop Hardware from the footer of every gotab.com page, and its product registry publishes 78 products with a price and a SKU code on every variant, across Terminals (8), Printers (11), Peripherals (26), Inventory (14), Bundles (4), Services (2) and Cables (1) - for example "Adyen AMS1" at 320.00, "Samsung A17" at 265.00, "Zebra DS2208 Barcode Scanner" at 225.00, "Epson TM-M30" at 310.00 and "GoTab On-Premise Server" at 500.00, with multi-variant bundles priced per configuration. Nothing at the SKU level is quote-only. Grade A because this is the publisher's own product registry rather than a marketing page. Three prior hosts named for this cell were all wrong - gotab.com/shop (the original no), store.gotab.com (the verifier's correction) and gotab.com/pricing, which carries no hardware SKU at all - so the analyst brief's claim that /pricing holds a hardware sheet is itself wrong. Shortfall to carry: install and setup fees remain quote-only, gotab.com/pricing answering that question with customised setup fees. Falsified if the storefront proves to be reseller- or login-gated rather than public. https://shop.gotab.io/products.json?limit=250 · retrieved 2026-09-05 adversarially verified

A
Yes

Hardware can be purchased outright (owned) as an alternative to a mandatory lease or subscription-bundled rental, and the vendor states which. differentiator

hardware-ownership-vs-lease

Both paths documented: outright purchase from GoTab (covered by the limited warranty) and Hardware Flex Leasing with $0 upfront and 36-month accruing credits toward buyout. Operators may also use their own devices. https://gotab.com/gotab-limited-warranty · retrieved 2026-08-01

B
Partial

Vendor documentation confirms whether hardware purchased from the vendor remains usable with other software after cancellation, i.e. it is not software-locked or remotely bricked. differentiator

hardware-usable-after-churn

Documentation settles the leased path, and settles it against the operator: clause 5.2 grants only a right to use the Equipment "solely in connection with the GoTab platform", and only "During the Subscription Term and while Customer is in good standing", while 5.1 provides that "All Equipment remains the sole property of GoTab unless and until the Hardware Balance is paid in full". Leased hardware is therefore contractually locked to GoTab and returnable. For hardware bought outright nothing addresses post-cancellation use with other software: the limited warranty and the merchant T&C return policy are silent, and the pricing-page line about not requiring exclusive use of GoTab hardware is about bringing your own devices, not about GoTab-sold devices working elsewhere. Note this addendum is incorporated into a "Master Subscription Agreement" GoTab does not publish (gotab.com/msa 404s), so the contract set is not closed. Falsified by an unlock or de-provisioning policy for purchased devices. https://gotab.com/hardware-flex-leasing-terms-and-conditions · retrieved 2026-09-05 adversarially verified

B
Partial

Staff can run the ordering or payment app on their own personal phone (BYOD) with a documented permission and security model.

hardware-byod

Software runs on commodity phones and tablets an operator already owns, which permits BYOD in practice; no documented BYOD permission or device-security model. https://gotab.com/faq · retrieved 2026-08-01

E
Yes

Integrates with specialty beverage hardware — self-pour tap walls, flow meters, or liquor pour spouts — such that pours post to a POS tab automatically.

hardware-selfpour-scales

Documented self-pour and beverage-hardware integrations: iPourIt, PourMyBeer, NAPA Technology, Pourtek, Wine Emotion, Smart Bar USA, BarTrack, Draft Control — and RFID Pass ties self-served pours to the guest's tab. https://gotab.com/products/rfid-pass · retrieved 2026-08-01

C

Integrations, API & extensibility

Yes

Public, self-serve API reference documentation is readable without a signed partner agreement, NDA, or sales call.

extensibility-public-api-docs

gotab.readme.io is fully public: guides, REST reference, GraphQL Explorer, changelog. No NDA, partner agreement or sales call required to read it. https://gotab.readme.io/ · retrieved 2026-08-01

B
Partial

API access is included in the base subscription at no additional per-location fee, surcharge, or upgraded plan requirement. differentiator

extensibility-api-access-cost

No API fee appears on the published price sheet and GoTab markets an Open API, but inclusion in the base subscription is not affirmatively stated. https://gotab.com/compare/toast-vs-gotab · retrieved 2026-08-01

D
Partial

A free sandbox or test environment with seeded data is available to developers without owning a paid production account. differentiator

extensibility-free-sandbox

The Quick Start guide states: 'If you are not already a GoTab API User or do not have a sandbox account please contact api.support@gotab.io', and separately 'If your application needs to test credit card authorizations notify api.support@gotab.io'. Sandbox access is gated behind a support email, not self-serve, and nothing states it is free. This also falsifies the dossier's note on reporting-public-api ('self-serve credential creation and a sandbox') and its structural note claiming the docs are 'self-serve (no NDA, no sales call)' — the docs are self-serve to read, the sandbox is not self-se https://gotab.readme.io/docs/quick-start · retrieved 2026-08-01 adversarially verified

B
Yes

Third-party apps authenticate via OAuth 2.0 with scoped, operator-granted, individually revocable permissions (not shared static API keys or credential sharing).

extensibility-oauth-partner-apps

Both OAuth 2.0 Client Credentials Grant and Authorization Code Grant Flow are documented guides, with Generate/Refresh Access Token endpoints. Scope granularity and per-operator revocation UI are not detailed. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Partial

The platform pushes real-time webhooks for order lifecycle events (created, modified, paid, voided, refunded) rather than requiring polling.

extensibility-webhooks-push

Push exists, under the name events-driven API. GoTab requires from the integrator a 'URL for GoTab to post loyalty events to (e.g. myloyaltyplatform.com/gotab/events)' and posts four event types to it: 'Customer driven, sychronous events: INQUIRE and REDEEM' and 'Asynchronous events: ACCRUAL, REVERSE'. ACCRUAL is 'an asynchronous request that will be triggered for each tab at a location as they are closed and updated in GoTab' and is sent 'for every closed tab', carrying the full tab payload of orders, items, adjustments and payments. The identical mechanism is documented for promo codes at /reference/gotab-promo-api.md. Partial rather than yes because the fired events are loyalty and promo lifecycle, not the order lifecycle the claim names: there is no created, modified, voided or refunded push, the closed-tab push is the only order-state event, and the endpoint is configured per integration partner rather than subscribed by the operator. Falsified by a readme page documenting a general order-event subscription, or downward if ACCRUAL proves to be batch rather than per-tab. https://gotab.readme.io/reference/gotab-loyalty-api.md · retrieved 2026-09-05 adversarially verified

B
No

Webhooks are cryptographically signed (e.g. HMAC) AND documented to retry with backoff, with a replayable event log for missed deliveries. differentiator

extensibility-webhook-reliability

The document that defines the GoTab outbound push also enumerates what the integrator must supply, and the security half is a shared secret rather than a signature: 'Authorization header value for your platform to confirm requests are coming from GoTab', which the sibling promo API marks as optional. No HMAC, signing secret, timestamp or signature header appears anywhere in the 102 readme pages. Reliability is equally undocumented: the word retry does not occur in this document, and corpus-wide it appears only in the two OAuth flow pages and in /reference/basic-accounting-integration.md, where the advice runs the other way and the integrator is told to write retry logic when polling fiscal-day status. The tokens Retry-After, idempotenc and webhook each return ZERO across all 877 documents on all five surfaces, and no replayable event log or delivery-history endpoint is documented. Falsified by any readme or changelog page naming a signature header or a delivery-retry policy. https://gotab.readme.io/reference/gotab-loyalty-api.md · retrieved 2026-09-05 adversarially verified

B
Yes

A public write API accepts externally originated orders into the POS as first-class tickets that fire to KDS/printers and land in reporting.

extensibility-order-injection-api

Public write endpoints: Create a tab, Add items to tab, Add items to tab by pass, Add payments to tab, Add Open Discount, Price Check, Cancel tab. Externally originated orders become first-class tabs. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
No

The vendor is a named DoorDash Preferred Integration Partner (DPIP) under the 2026 program requirements. differentiator

extensibility-doordash-preferred

DoorDash publishes its own complete provider registry at https://merchants.doordash.com/api/partner-integrations — 64 providers, each carrying a `qualityStatus` of PREFERRED (12), CERTIFIED (47) or AT_RISK_OF_SUSPENSION (5), plus a count of how many of the eight high-quality features the provider has built. It is served anonymously and merchants.doordash.com/robots.txt permits it (Allow: / with only /cdn-cgi/ and /monitoring barred). This is what the program page means when it says the roster is tracked in real time, and it is a CENSUS rather than a floor: every provider integrated with DoorDash appears in it with a tier, so a vendor that is absent altogether is not in the programme at any tier. This vendor appears nowhere in the 64 entries, under any provider id or friendly name — checked mechanically against both fields, with the only near-misses being substring coincidences (Speedline against Lightspeed, vroom_restaurant against Restaurant365) that were read and rejected. SUPERSEDES the 2026 announcement post most of this corpus cited: that post introduces its ten names with the word `include`, which makes it a floor and never a ceiling, and reading it as a ceiling is what produced this cell's prior grade. GRADE HARMONISED TO B, 2026-09-11: this URL was first written at grade A. The adversarial pass ruled E on the principle that the rubric grades by the publisher's relationship to the SCORED vendor, and DoorDash is not that vendor. B is the settlement, for two reasons that are not preference: 39 rows already graded a DoorDash-published document B on this claim before today, and one URL on one claim cannot carry two grades; and E is unworkable, because nine of the 47 rows are `yes` on a differentiator-weight claim, which the grade floor refuses below B. B reads as the product documentation of the programme's own publisher, which is what this registry is. The dissent is recorded in verification.jsonl. https://merchants.doordash.com/api/partner-integrations · retrieved 2026-09-11

B
Partial

Direct, first-party certified integrations exist with DoorDash, Uber Eats, and Grubhub without requiring a paid middleware layer (Otter/Chowly/Deliverect/Checkmate). differentiator

extensibility-first-party-delivery-integrations

Uber Eats, Deliveroo and Just Eat are named direct partners. DoorDash and Grubhub are not in the directory, so the DoorDash/Uber Eats/Grubhub trio is not covered first-party. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

The POS is a supported endpoint for at least two major aggregation middleware platforms (Deliverect, Chowly, Otter, Checkmate, ItsaCheckmate).

extensibility-middleware-compatibility

GoTab is a supported endpoint for exactly one of the five platforms the claim enumerates, against a bar of two. OTTER, verified on the integrator's own site (retrieved 2026-09-15): tryotter.com/integrations lists GoTab under the category heading 'Point of Sale' — not 'Online ordering', where that same index files Flipdish, Slice, ChowNow and Menufy by HungerRush — and the GoTab page is titled 'Integrate with GoTab | Extend the power of your POS system with Otter', saying 'With Otter's GoTab integration, you can send all your delivery orders directly to your POS – no manual labor required', 'Automatically accept every order that comes in and send it straight to GoTab', and 'Send order to your GoTab POS — See all of your orders in one place and send them directly to your GoTab POS system.' Direction therefore runs INTO the GoTab POS. GoTab's own documented aggregator route is Stream, which is NOT one of the five: the 3rd Party Ordering help article (docs.gotab.io, grade B) has the operator create a 'managerial Stream User ... the user that keeps your location connected to the 3rd Party Ordering platform', authorize at gotab.streamorders.com/welcome, after which 'your DoorDash, UberEats etc. orders are facilitated through the 3rd Party Ordering platform into GoTab. From GoTab, you can view the orders in your POS and fulfill the orders on your KDS.' SHORTFALL: one enumerated platform, not two; and the only grade-B document in the cell evidences Stream, an off-list aggregator, so the enumerated leg rests entirely on Otter's own marketing page — hence grade C. DELIVERECT MUST NOT BE COUNTED: its GoTab page runs the other way, labelling the entry 'Category Online ordering' and stating 'Orders made on GoTab are then integrated directly to your POS', i.e. it treats GoTab as an ordering CHANNEL feeding someone else's POS. ItsaCheckmate's integrations page enumerates 28 POS systems and GoTab is not among them (0 occurrences in 153 KB). Chowly is a robots-barred host and was not measured. CONTROLS: tokenised on word boundaries over GoTab's own 560-file corpus, `deliverect`, `chowly`, `otter`, `itsacheckmate`, `checkmate`, `cuboh` and `ordermark` each return 0 while positive controls fire (doordash 27, ubereats 36, grubhub 16, streamorders 9) — GoTab's own documentation never names any of the five. Substring poison is real here: `stream` is `streamline` 17 times in this corpus, and raw `otter` substrings run 1486 against 147 tokenised. Soft-200 control: tryotter.com/integrations/zzqxnonsense returns HTTP 404, so the GoTab page is a real route and not a catch-all. Overturned to yes by any second enumerated platform documenting GoTab as a POS endpoint. https://www.tryotter.com/integrations/gotab · retrieved 2026-09-15 adversarially verified

C
Yes

Native, vendor-maintained connectors exist for QuickBooks Online and at least one other GL/accounting system, with mapped journal entries rather than CSV-only export.

extensibility-accounting-connectors

QuickBooks Online sync described as native (at no extra cost per the vendor's comparison page, though the price sheet lists accounting integration at +$10/mo on Basic), plus Restaurant365, Shogo, DAVO, Compeat and Buyers Edge Back Office. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

Labor hours export natively to at least two named payroll providers (e.g. Gusto, ADP, Paychex, Paylocity) without manual re-keying.

extensibility-payroll-export

Labor/tip integrations with Workforce.com, 7shifts, Schedulefly, TipHaus and eTip, but no named major payroll provider (Gusto, ADP, Paychex, Paylocity). https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

The vendor operates a public integrations marketplace listing named third-party apps that operators can browse and self-install.

extensibility-app-marketplace

A public, browsable integration directory lists roughly 100 named partners across 20+ categories, and there is a Marketplace product page — but self-install by the operator is not documented. https://gotab.com/integration-partners · retrieved 2026-08-01

B
Partial

A complete historical data export (orders, customers, menu, payments metadata) is available to the operator on demand, in a documented machine-readable format, at contract end. differentiator

extensibility-data-portability-exit

Broad read APIs make a technical export possible on demand, but there is no documented machine-readable full-history export product and the merchant T&C specifies no contract-end export process. https://gotab.com/merchant-terms-and-conditions · retrieved 2026-08-01

B

Reliability, offline & operations

Yes

Order entry, ticket routing, and check printing continue on-terminal when the internet connection is lost.

reliability-offline-order-entry

Order entry, printer routing and KDS continue on-premises via OPServer during an outage. https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-08-01

C
Partial

Card payments can be accepted while offline via store-and-forward, not just cash-only fallback. differentiator

reliability-offline-card-auth

Three problems. (1) The sole sources are a marketing product page and the FAQ, not documentation. (2) The capability is gated: OPServer is listed on the pricing page as 'Not available' on Basic, +$75/mo on Pro, included only on Sync $229 — this is a paid add-on on a mid tier, not a platform property, and the dossier's own pricing block omits the Basic exclusion. (3) The mechanism is 'GoTab's standalone payment terminals continue processing credit and debit card payments offline' — 'standalone' is the Adyen term of art for a terminal operating detached from the POS, and Adyen's own docs describ https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-08-01 adversarially verified

C
No

The vendor publicly documents who bears the loss when a store-and-forward transaction declines after connectivity returns, and states any per-transaction or cumulative offline cap. differentiator

reliability-offline-decline-liability

The OPServer page is GoTab's own itemised account of offline behaviour and it fills every column except this one. It affirms offline card acceptance - "standalone payment terminals continue processing credit and debit card payments offline" and "cash transactions are always an option" - and states "Transactions are stored securely until your connection returns", with reconnect handled by "GoTab automatically syncs all transaction data, order history, and inventory updates back to the cloud. No data loss, no extra steps". It names the exclusions it does have (QR ordering requires an internet connection; third-party labor systems require an internet connection) yet never states a per-transaction or cumulative offline ceiling, a time limit, or who absorbs a decline discovered at settlement. Searched all five surfaces: 'declin' hits 4 of 877 docs and none is a GoTab liability statement, the Merchant Terms and Conditions allocate no offline risk, and the KB has no offline-payment article. Falsified by any published offline floor limit or decline-responsibility sentence. https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-09-05 adversarially verified

C
Partial

Multiple terminals continue to share a single check/table state over the local network while the internet is down (LAN-only degraded operation, not per-terminal islands). differentiator

reliability-lan-degraded-multi-terminal

P4 is right: the product page, not the KB, is where GoTab describes this, and it asserts shared state three times. "A local server keeps every device" - POS, kitchen, kiosks, payments - "talking to each other and taking orders even when the internet drops"; "All POS terminals, handhelds, kiosks, and payment devices use a shared tab model. They all stay connected to each other, even when you're offline."; and the outage FAQ answer "Your full shared tab model remains active on-premises. Whether guests open a tab via handheld or fixed POS devices, kiosks, or payment terminals, everything syncs locally, so there". That is LAN-shared check state rather than per-terminal islands, stated by the vendor. Held to partial for three reasons: the source is a marketing product page with no architecture, sync scope or device-count limits (grade C); OPServer is a paid add-on (gotab.com/pricing: not available on Basic, +$75/mo on Pro, included on Sync), so this is not a base-product property; and without it the documented behaviour is the opposite, the 5/9/25 POS release note describing one terminal buffering its own in-progress order and sending it after reconnection. Searched all five surfaces. Falsified by a spec page showing per-terminal isolation under OPServer. https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-09-05 adversarially verified

C
Partial

The vendor documents a local/on-premise transaction engine or edge server (not purely cloud-hosted request/response) for the ordering path. differentiator

reliability-local-transaction-engine

What is documented: GoTops, a locally installed 'POS and KDS app' for iOS, Android, Windows and Mac, published so operators need 'never worry about how to use GoTab during an internet outage'. Named shortfall: that is a locally installed client, not a documented on-premise transaction engine for the ordering path. The named server product - 'On-Premise Server Connection', priced on gotab.com/pricing as not available on Basic, +$75/mo on Pro and included on Sync $229/mo, and marketed as OPServer - is described nowhere in GoTab's documentation: the help-centre search API returns zero articles for 'OPServer' and docs.gotab.io (guides, REST and GraphQL references) has no on-premise material. Its architecture, sync model and limits are unpublished. https://success.gotab.io/knowledge/downloading-and-using-the-gotops-app · retrieved 2026-08-06

B
Yes

Kitchen display and/or kitchen printer routing continues to function while offline.

reliability-offline-kds-printing

KDS continues operating without interruption and orders keep routing to printers offline. https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-08-01

C
No

The system automatically fails over to a backup printer or KDS station when a configured kitchen printer is unreachable, and alerts staff.

reliability-printer-fallback

The KDS settings overview enumerates the complete Devices & Printers configuration — Fulfillment Printer, Closeout Printer, Pair to TDS — with no backup-printer or failover setting, and the POS notification set covers product availability, refund requests, QR ratings and QR scans only, so no staff alert fires when a printer is unreachable. The printer-down support article confirms the behaviour: jobs queue as 'pending print jobs' while staff manually power-cycle the printer, switch and GoTab box. https://success.gotab.io/knowledge/kds-settings · retrieved 2026-08-04

B
No

The vendor documents its conflict-resolution behavior for edits made on two devices during a network partition (e.g. last-write-wins, merge, or operator prompt).

reliability-sync-conflict-handling

GoTab publishes an outcome and no mechanism. The reconnect FAQ says "GoTab automatically syncs all transaction data, order history, and inventory updates back to the cloud. No data loss, no extra steps", and the body repeats "When your internet or cloud connection comes back, all locally stored data syncs back to the cloud. This happens automatically and securely." Nowhere is there a rule for two devices editing the same tab across a partition: no last-write-wins, no merge, no operator prompt. Searched all five surfaces - 'conflict|last.write|merge conflict|concurrent' hits 14 of 877 docs, all of them cash reconciliation, menu scheduling, packing slips, preauthorizations or legal conflict-of-law boilerplate, and the readme REST/GraphQL references expose no concurrency contract ('etag|if.match|optimistic|409' hits three docs, none of them a version-token spec). This is a documentation-existence claim and the documentation does not exist. Falsified by any published conflict rule or an API concurrency token. https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-09-05 adversarially verified

C
Partial

The vendor publishes an explicit list of which features are unavailable offline (gift cards, loyalty lookup, refunds, manual card entry, text-to-pay).

reliability-offline-feature-matrix

Vendor affirmatively names QR ordering, third-party labor systems and DoorDash/Uber Eats as internet-required, and names order entry, printing, KDS, card terminals, cash, inventory and clock-ins as working. Gift cards, loyalty lookup, refunds and manual card entry are not addressed. https://gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-08-01

C
Yes

A public, per-component status page exists showing current and historical incident state without a login.

reliability-public-status-page

Verified myself and the dossier understated nothing: login-free, per-component current status plus rolling 90-day uptime (Infrastructure 99.99%, POS 99.98%, KDS 99.98%, CFD 99.99%, QR Ordering/Kiosk/Print/Manager Dashboard/SMS/RFID/Payments 100.0%), dated incident history including the 24 Jul 2026 POS-CFD connection issue and a live degraded Twilio SMS dependency, with third parties tracked separately. https://status.gotab.io/ · retrieved 2026-08-01 adversarially verified

B
Partial

24/7/365 live human support by phone is included in the base subscription, not sold as a premium support tier.

reliability-247-live-support

24x7x365 support via email, phone, chat and in-app exists — but it is a +$35/mo add-on on the Basic tier, included only in Pro and Sync. Not in the base subscription. https://gotab.com/pricing · retrieved 2026-08-01

C
Partial

The vendor offers in-person on-site installation and go-live support (first-party or certified dealer network), not remote-only self-install. differentiator

reliability-onsite-install

On-site presence is offered but the documented default is remote. The homepage support block promises "white-glove treatment from onboarding and hardware deployment, to go-live and account management" with a "Dedicated onboarding/project management team", and lists "On-site training availability" as one training option - but the enumerated onboarding steps end at "Finalizing on-site lab setup for initial remote training", i.e. a lab built so that training can be delivered remotely. No first-party installation service, no certified dealer network and no cutover-night attendance is described anywhere on the four surfaces I searched (gotab.com, success.gotab.io, gotab.readme.io, opsi.io; status.gotab.io not fetched). Corroborating but weaker, and in a different document: gotab.com/gotab-limited-warranty voids coverage for installation not completed by a GoTab representative, which presupposes that GoTab representatives do install. Falsified by a published implementation-services or dealer page. https://gotab.com · retrieved 2026-09-05 adversarially verified

C
No

The vendor performs the initial menu build as part of onboarding rather than leaving it entirely to the operator. differentiator

reliability-menu-build-service

The homepage enumerates what GoTab's Customer Success team does during onboarding - "Identifying all major project needs in a Scope of Work", documenting bespoke feature requests and integrations, and "Finalizing on-site lab setup for initial remote training" - and menu build is not among them. Corpus-wide 'menu build|build your menu|menu setup|configure your menu|menu import|migrate your menu' returns 1 document across all four surfaces, and the KB documents menu creation as an operator task in the manager dashboard (success.gotab.io/knowledge/getting-started-creating-a-menu). Two things keep my confidence low and I flag rather than resolve them: gotab.com/compare/clover-vs-gotab asserts that GoTab's onboarding team handles setup and staff training as part of the transition without saying what 'setup' covers, and merchant T&C section 6 makes configuration services purchasable under a separate SOW at an unpriced setup fee. Falsified by either being pinned to menu entry. https://gotab.com · retrieved 2026-09-05 adversarially verified

C
Partial

A hardware replacement/advance-exchange program with a stated turnaround time is included or purchasable.

reliability-hardware-replacement-sla

A warranty/RMA program exists with published coverage terms, but the remedy is a resolution within a reasonable time — no stated turnaround. https://gotab.com/gotab-limited-warranty · retrieved 2026-08-01

B
No

The vendor publishes a current PCI DSS v4.0/4.0.1 Attestation of Compliance or equivalent (e.g. validated P2PE listing) covering the post-March-2025 future-dated requirements.

reliability-pci-dss-4-attestation

MERGE: grade B -> C, the cited page is a marketing blog post (marketing-cited no = C). 328th, THE PUBLICATION CENSUS THIS CLAIM NEEDS IS NOW COMPLETE, and it is negative. The claim is a publication claim ('the vendor publishes a current PCI DSS v4.0/4.0.1 AoC or equivalent'), so a complete census of what GoTab publishes settles it. NEW SURFACE: gotab.com was rebuilt (Webflow, 'Last Published: Thu Sep 17 2026') and its sitemap - which prior passes sampled at 422 of ~3,275 URLs - was fetched IN FULL this pass: /sitemap.xml lists 3,254 locs, of which 1,601 are the /es Spanish mirror of the same pages; all 1,653 English URLs were fetched, 139,124,889 bytes, and ZERO of them returned the site's constant 15,688-byte 404 body (measured), so the corpus is the whole published English site. TOKENISED CONTROLS over all 1,653 pages: P2PE 0, 'point-to-point' 2 (both POISON - a legacy-vs-cloud blog table using 'point-to-point' to mean bespoke integrations), AoC 0, '4.0.1' 0, 'Level 1 Service Provider' 0, 'trust cent(er|re)' 0, 'SOC 2' 0, QSA 1, ROC 1, SAQ 4. `attestation` returns 4 hits and ALL FOUR are in one blog explainer, gotab.com/latest/a-guide-to-pci-compliance, describing what MERCHANTS at each PCI level must do ('Complete the Attestation of Compliance Form'); the only self-statement on that page is 'GoTab is proud to announce that we are Level One Merchants in the PCI Compliance level' - a merchant level, not a service-provider validation, with no DSS version, no QSA, no validation date and no document. The 11 'PCI DSS' hits are generic prose (an RFP checklist, 'PCI DSS v4.0 requirements accelerating modernization timelines', and the FAQ's 'GoTab's iPad POS system uses the same PCI-compliant payment processing standards as traditional hardware'). ALSO NEGATIVE AND NEW: the two merchant agreements the rebuilt site publishes at /merchant-terms-and-conditions (110,343 b) and /australia-merchant-terms-and-conditions (111,717 b) - documents no prior pass had found, because /terms, /msa and /legal all return the 404 body - contain PCI 0, P2PE 0, 'Payment Card Industry' 0, attestation 0, encrypt 0. BOUND, stated in full: this census (1,653 pages) plus the previously-censused docs.gotab.io (335 pages via its llms.txt), the success.gotab.io knowledge base (278 sitemap URLs, bodies read), all 37 readable Partner Resource Center PDFs (aoc 0, attestation 0, p2pe 0, qsa 0) and the shop.gotab.io hardware catalogue (78 products, the only PCI token being 'PCI PTS 6' on a Verifone P630, a device approval); and crt.sh's 94 gotab.io/gotab.com hostnames contain no trust., security. or compliance. subdomain, so there is no unlinked trust portal. GoTab publishes no v4.x Attestation of Compliance and no validated P2PE listing anywhere. (The /es mirror is a translation of these same pages and was excluded deliberately.) Retrieved 2026-09-18. https://gotab.com/latest/a-guide-to-pci-compliance · retrieved 2026-09-18 adversarially verified

C
Partial

Multi-factor authentication is enforceable on back-office and POS administrative logins, with role-based permissions.

reliability-mfa-role-based-access

Role-based access is documented via User Roles APIs, and OTP challenge/verify endpoints exist in the public API — but MFA enforcement on back-office and POS administrative logins is not documented. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Partial

A public, free training library (video/LMS) and a practice/training mode on the terminal exist for staff onboarding.

reliability-self-serve-training

Public guides, video and podcast libraries plus Partner Training and a Partner Resource Center exist on gotab.com, and a sandbox exists for developers; a free staff LMS and a terminal training mode are not documented. https://gotab.com/partner-training · retrieved 2026-08-01

B
No

Any terminal can assume the master/server role automatically if the designated primary terminal or local server fails. differentiator

reliability-failover-terminal-role

OWNER RULING 2026-09-18 (Q5): a cloud-hosted POS with no in-store master/primary terminal role scores `no` on reliability-failover-terminal-role rather than unknown. Landed no/B on the evidence below, which establishes that architecture. Retrieved 2026-09-18. PRIOR EVIDENCE: CLAIM: any terminal assumes the master/server role if the primary terminal or local server fails. VERDICT: stays unknown. NEW SURFACE: crt.sh (%.gotab.io 92 names, %.gotab.com 10) gave 94 unique hosts; probed root, robots.txt and /swagger, /swagger/index.html, /swagger/v1/swagger.json, /api-docs, /openapi.json, /docs, /redoc, /graphql, /sitemap.xml, /llms.txt, /v3/api-docs on the live ones. No new anonymous documentation surface: api/beta/staging/pos/kds.gotab.io are the same 16.7 KB SPA shell (all spec paths 302 to /not_found); mcp2.gotab.io and datahub.gotab.io 401; ecomm-store-agent.gotab.io 401 "Missing required headers"; deprecated.docs.gotab.io, discover.gotab.io, es.gotab.io/com are Cloudflare 403 "CNAME Cross-User Banned" (refusal, not absence); sdk.gotab.io 404; docs.gotab.io/llms.txt lists 335 pages, identical to the q310 corpus except graphql-schema.txt and specs/gotab-api.json (both already censused). The one new first-party surface is shop.gotab.io, a Shopify hardware store whose products.json (limit=250; page 2 empty) is a complete 78-product catalogue. The On-Premise Server listing: "GoTab On-Premise Server includes : 1x - GoTab On-Premise Server 1x - power cable 1x - RJ45 cat 6 ethernet cable What is a GoTab On-Premise Server? On-premise server running the local GoTab devices, synced with the cloud in real-time when connected to ISP and continues to runs the network locally in the case of an internet outage" - a single-box topology whose continuity statement is again about ISP loss, not server loss. Every printer listing adds "*Please note for printers, a GoTab Print Server is required at your location". The only 'failover' tokens (2 of 78 products) are the Secure Gateway UCG/UDM: "dual WAN supporting internet failover" / "multiple WAN's supporting internet failover" - network WAN failover, a homonym for this claim. Controls over the 78 bodies: redundant 0, backup 0, offline 0, outage 1 (the server sentence). No terminal-role enumeration, so no no; nothing claims promotion, so no yes. Retrieved 2026-09-16. || 322nd: VERDICT: unknown (held). NEW SURFACE (2026-09-17): the display boot app https://entry.gotab.io/ ("GoTab -- Display Entrypoint", production; Vite, /.vite/manifest.json lists exactly two entries, index.html -> assets/main-C6_Eml0Y.js (1,671 bytes) and debug.html; no public source map, the .map URL returns the SPA shell). Its whole logic is readable: it hard-codes local origins "https://tunnel.gotablocal.com" then "https://home.gotablocal.com,https://home1.gotablocal.com", probes each `${origin}/metadata` with a 2.5 s timeout, accepts the first whose locations include `on_premise_preferred` ("Found local server at ${o}"), and otherwise navigates to the cloud "https://gotab.io/displays". So a display prefers an on-premise GoTab server when one answers and FALLS BACK TO THE CLOUD; two local hostnames (home, home1) exist, but nothing in the code promotes a terminal to the server role. CT listing for %.gotablocal.com (crt.sh, 123 names) is per-merchant local-server certificates (108 m-<id>-<hash>-<venue> names, plus micro-* and home/home1/tunnel); home.gotablocal.com resolves to private 10.95.88.16 and home1 to 127.0.0.1, so these are LAN endpoints, not a public documentation surface. The success.gotab.io KB (sitemap 278 URLs, bodies read) has 0 hits for on-premise / local server / failover / backup server; its GoTops article says only "Never worry about how to use GoTab during an internet outage". Evidence describes cloud fallback and ISP-outage continuity, not terminal promotion on server failure: no yes; no enumeration of roles, so no no. Retrieved 2026-09-17. || 327th: NEW MEASUREMENT (327th), cell unchanged. Two negatives worth recording so a later pass does not repeat them. (1) GoTab's complete published REST API (docs.gotab.io/specs/gotab-api.json, OpenAPI 3.1, 65 paths / 90 operations / 72 schemas, production server https://gotab.io/api) has no device, terminal-role, primary/secondary, election or failover concept of any kind; the only terminal object is the payment terminal (/v2/loc/{location}/payment-terminals and its checkouts), which is a card-reader resource, not a POS role. (2) pos.gotab.io and kds.gotab.io are NOT separate applications - both serve the byte-identical 16,699-byte shell whose meta description is 'Mobile payment, ordering, and rewards' and whose 59 preloaded scripts and 27 stylesheets all come from cdn.gotab.io/static, i.e. they are aliases of the GUEST ordering web app, not store-level POS or KDS builds. That app's webpack runtime (cdn.gotab.io/static/uuid_1-yi~d044e90c.c38b77ad.js) publishes its whole chunk table - 37 eagerly installed chunk ids plus exactly 21 lazily loadable ones (ids 21-33 and 2032-2041, with their content hashes and a parallel CSS hash map) at publicPath https://cdn.gotab.io/static/ - so the guest app IS bounded, but it is the wrong app for this claim and no store-level GoTab POS bundle is anonymously reachable. Retrieved 2026-09-18. || 328th: 328th, NEW SURFACE AND THE MOST EXPLICIT ARCHITECTURE STATEMENT GOTAB HAS EVER PUBLISHED - it still does not carry the claim. The rebuilt gotab.com ships a dedicated OPServer product page, /products/on-premise-server-with-enhanced-offline-mode, which no prior pass could have seen. Verbatim: 'GoTab's OPServer is a small local server, installed on-site, that every GoTab device on your network connects to - point of sale, kitchen displays, kiosks, handhelds, and printers. It's always running the show.'; 'Every device stays in sync, not just one terminal - POS terminals, kitchen displays, kiosks, handhelds, and printers all keep communicating with each other through OPServer'; 'While the internet is down, OPServer assumes everything it last knew to be true still is - menu availability, 86'd items, pricing rules, membership and cart rules'; 'Cards are captured and held on the payment terminal or OPServer until the connection returns'; 'QR and mobile ordering, and text messages, do need the internet'. The matching FAQ entry: 'The moment GoTab detects an internet disruption, it automatically switches to on-premises mode.' EVERY ONE OF THESE IS ISP FAILOVER TO A SINGLE DEDICATED BOX, which is the opposite direction from the claim: the claim asks what happens when that box (or the primary terminal) FAILS. CONTROLS over the complete 1,653-page English gotab.com census (139,124,889 bytes): 'primary server' 0, 'primary terminal' 0, 'secondary server' 0, 'redundant server' 0, 'promote to' 0; `master` 86 hits, ALL POISON (a 'Master Sommelier' blog post repeated in its own meta tags, 'Master Printing', 'Master Subscription Agreement', 'master their shift', 'Mastercard'); `failover` 10 hits, ALL POISON or off-target - 6 are the Australian merchant agreement's optional 'failover' PAYMENT PROCESSOR clause, 1 is that agreement's data-backup covenant, and the other 3 are blog prose about the cloud/offline mode ('No manual failover', 'automatic local failover for supported venue operations') plus a partner's 'cellular failover'. So across the vendor's whole published site there is no terminal-role, election, promotion or server-redundancy concept at all - but a marketing site is not an enumeration of terminal roles, so this is not published as a `no`. Retrieved 2026-09-18. || 330th: MERGE HOLD (proposed no/B, held unknown): OWNER QUESTION 5 (is 'cloud POS with no master terminal' a `no` for this claim?) is undecided; the row is ready and lands no/B if the owner says yes. Evidence kept: KB fixpoint 304 pages with opserver/on-premise/local server/failover/offline mode all 0, and entry.gotab.io falls back UP to cloud, never promoting a terminal. PROPOSAL AS WRITTEN: 330th: PROPOSED MOVE to `no`, with the owner's pending question answered as directly as the evidence allows. NEW MEASUREMENT: the operator-facing KB was re-derived to an href fixpoint (304 pages, 825,520 chars - the 278-URL sitemap plus 26 category pages linked only from article bodies) and searched with GoTab's own vocabulary: `opserver` 0, `on.premise` 0, `local server` 0, `failover` 0, `offline mode` 0. The single continuity sentence in the entire KB is in 'Downloading and Using the GoTops App': 'Never worry about how to use GoTab during an internet outage.' So the manuals an operator is given contain no server, no primary/secondary terminal, and no role concept at all. Set against the three surfaces already on record, the architecture is now described from every side and none of them is terminal promotion: (1) the OPServer product page says one box is 'always running the show' and that 'every GoTab device on your network connects to' it - a single dedicated appliance, not an elected terminal; (2) every published failure story is ISP loss, not box loss ('The moment GoTab detects an internet disruption, it automatically switches to on-premises mode'); (3) the one piece of GoTab code that handles a missing local server - the display boot app entry.gotab.io, whose whole logic is readable - probes tunnel/home/home1.gotablocal.com for 2.5 s each and, if none answers, navigates to the CLOUD (https://gotab.io/displays). That is a documented automatic fallback in the opposite direction: devices fall back UP to the cloud, they are never promoted INTO the server role. Controls over the complete 1,653-page site census remain 'primary server' 0, 'primary terminal' 0, 'secondary server' 0, 'redundant server' 0, 'promote to' 0, with `master` 86/86 poison (Master Sommelier, Mastercard, Master Subscription Agreement) and `failover` 10/10 poison (a payment-processor failover clause, cellular/WAN failover). WHY THIS IS NOW A `no` AND NOT A HOLD: GoTab is a full POS whose own docs, site, published API and boot code are all censused, and the merge policy allows a `no` for a module a full POS lacks once its own docs are bounded; the 'claim does not apply, hold unknown' carve-out was written for middleware and voice products. The honest caveat for the owner: this is a `no` on the mechanism (no terminal assumes the master/server role), not a statement that a venue goes down - the published behaviour on local-server loss is cloud fallback, which is why a reader might prefer `n-a`. Retrieved 2026-09-18. https://www.gotab.com/products/on-premise-server-with-enhanced-offline-mode · retrieved 2026-09-18 adversarially verified

B
Partial

Terminals support an automatic cellular/LTE failover for connectivity, first-party or documented.

reliability-cellular-backup

Cellular connectivity is documented for exactly one device class: the Samsung A17 handheld can 'use an optional SIM card for off-site events'. Shortfalls: positioned as off-site connectivity rather than an automatic LTE failover for the venue, no failover behavior is documented, and no cellular path exists in the docs for fixed terminals, KDS or payment devices — GoTab's documented outage answer is the OPServer offline mode, and its FAQ instead addresses locations with 'bad cellular service and/or bad WiFi'. https://gotab.com/products/pocket-phone-only-pos · retrieved 2026-08-05

C

Commercial, compliance & data ownership

Partial

Vendor publicly offers a month-to-month subscription with no minimum multi-year term commitment for its core POS software tier. differentiator

commercial-month-to-month-contract

Vendor contradicts itself: the pricing page says GoTab does not require long-term contracts, while its own Toast comparison page states 2-year standard terms. The T&C mechanism is a negotiated Initial Term on the Order Form, then genuine month-to-month renewal with 14 days notice. https://gotab.com/compare/toast-vs-gotab · retrieved 2026-08-01

B
Partial

Vendor's published terms of service or MSA state no early termination fee / no liquidated damages for cancelling the software subscription before term end. differentiator

commercial-no-early-termination-fee

No software ETF or liquidated-damages clause found in the merchant T&C. But the Hardware Flex Leasing addendum accelerates the entire unpaid Hardware Balance on termination for any reason, and returned equipment does not reduce the obligation. https://gotab.com/hardware-flex-leasing-terms-and-conditions · retrieved 2026-08-01

B
Yes

The auto-renewal term and the required cancellation notice window (e.g. 30/60/90 days) are stated in publicly accessible contract terms rather than only in a signed quote.

commercial-autorenew-terms-published

Confirmed in the T&C: initial period specified on the Merchant Order Form, automatic monthly renewal thereafter unless either party gives 14 days' written notice. Also confirmed the 90-day fee-modification clause, the 30-day cure-based termination, absence of any software ETF or liquidated damages, merchant retention of Merchant Content ownership, the Stripe/Paysafe/Adyen processor set, and the absence of any uptime SLA or service credits — the dossier's commercial reading is accurate throughout. https://gotab.com/merchant-terms-and-conditions · retrieved 2026-08-01 adversarially verified

B
Partial

The POS software can be purchased and operated with a third-party payment processor of the operator's choice (no mandatory in-house processing lock). differentiator

commercial-processing-not-bundled

Three processors are named in the T&C (Stripe, Paysafe, Adyen), so there is choice within GoTab's set — but the FAQ states GoTab processes payments and deposits funds, and card rates are published as GoTab's own. Not open bring-your-own-processor. https://gotab.com/merchant-terms-and-conditions · retrieved 2026-08-01

B
No

Vendor publishes an interchange-plus (cost-plus) processing rate option with the markup stated as a specific bps + per-transaction figure, rather than only blended/flat card-present rates. differentiator

commercial-interchange-plus-published

Only a blended card-present rate is published (starting at 2.40% + $0.15, +0.85% Amex, +1.5% international). No interchange-plus option with a bps + per-transaction markup appears anywhere public. https://gotab.com/pricing · retrieved 2026-08-01

C
No

The processing agreement caps or prohibits unilateral vendor-initiated rate increases outside of pass-through interchange/network changes, or grants penalty-free exit on any increase. differentiator

commercial-rate-increase-clause

The opposite: GoTab reserves the right to modify the Fees payable hereunder upon written notice to Merchant at least ninety (90) days prior to the end of the then-current term. No cap on increases and no explicit penalty-free-exit-on-increase right, though the 14-day non-renewal window is a practical out. https://gotab.com/merchant-terms-and-conditions · retrieved 2026-08-01

B
Yes

Per-location and per-additional-terminal software pricing is published as specific dollar figures on the vendor's public site (not 'contact sales' only).

commercial-pricing-published

Specific dollars published: Basic $15/mo, Pro $99/mo, Sync $229/mo, plus additional counter or handheld POS at +$25-$35/mo each and named per-module add-on prices. Among the most transparent price sheets in the category. https://gotab.com/pricing · retrieved 2026-08-01

C
Partial

Individual modules (online ordering, KDS, loyalty, inventory, payroll, marketing) can be purchased à la carte and cancelled independently without repricing the base POS subscription. differentiator

commercial-module-unbundling

Published a la carte ADD-ON pricing: KDS +$25/mo (Basic) or +$35/mo, Kiosk +$25/+$35, additional counter or handheld POS +$25/+$35, Easy Tab +$25 on Basic, Loyalty + Memberships +$35, Gift Cards +$35, Order Status SMS +$50, 3rd Party Aggregation +$55, Multi-location management +$50, On-Premise Server Connection +$75 on Pro, and Opsi invoice +$85 / recipe, inventory and task management +$45 each. Named shortfalls: (1) unbundling is tier-gated, not open - Loyalty, Multi-location, On-Premise Server, Fraud/Insufficient Funds Protection and Automatic Remittances are all 'Not available' on the $15 Basic plan, and 24x7 support is itself a +$35 add-on there; (2) nothing published states a module can be dropped independently without repricing the base plan. The Merchant Terms and Conditions set Fees by reference to the per-merchant Order Form and govern term, renewal (monthly after the initial term, 14 days notice of non-renewal) and termination only for the Agreement as a whole, with no per-module cancellation provision. https://gotab.com/pricing · retrieved 2026-08-06

C
Partial

All required terminals, KDS screens, and printers can be purchased outright at a published price, with no mandatory multi-year hardware lease or rental.

commercial-hardware-purchase-outright

Outright purchase from GoTab is supported (warranty covers hardware purchased directly from GoTab) and operators may use their own devices, so no lease is mandatory — but no hardware prices are published. https://gotab.com/gotab-limited-warranty · retrieved 2026-08-01

B
Yes

Vendor documents support for at least one non-proprietary/third-party hardware option (e.g. standard iPad, Epson/Star ESC/POS printers, generic cash drawer) rather than requiring vendor-only devices. differentiator

commercial-hardware-not-locked

We try to use all of your existing receipt printers, tablets, and KDS systems; software runs on commodity Android/iOS devices and Windows PCs. https://gotab.com/faq · retrieved 2026-08-01

B
Partial

Operators can self-serve export their full transactional history (orders, line items, payments, labor) to CSV or a documented API without contacting support or paying a fee.

commercial-data-export-self-serve

Self-serve API credentials plus REST/GraphQL reads across tabs, tab items, payments, customers and inventory make programmatic export possible without a support ticket; no documented CSV export of full transactional history in the UI. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Partial

Guest/customer records, loyalty balances or point ledgers, and gift-card liability balances are exportable in machine-readable form by the operator. differentiator

commercial-export-customer-and-loyalty

List customers and segment membership are API-exportable; loyalty point ledgers and gift-card liability balances are not documented as exportable objects. https://gotab.readme.io/llms.txt · retrieved 2026-08-01

B
Partial

Contract or documentation specifies a post-termination data retrieval window (a defined number of days during which the operator can still export their data) rather than immediate cutoff. differentiator

commercial-post-termination-export-window

A defined window exists, but only in the Australian terms. Clause 2.4 provides that following the Term GoTab "will use commercially reasonable efforts to make available access to the GoTab Solution to enable Merchant to download reports and export extracts of Merchant Content", and that "GoTab may charge reasonable rates" where such access is made available "more than 60 days after the end of the Term" - a free 60-day retrieval period, chargeable thereafter. The US merchant terms at gotab.com/merchant-terms-and-conditions have no counterpart: section 11.3 provides that on termination all licenses granted thereunder immediately terminate, and section 5.2 states that GoTab is not obligated to back up any Merchant Content (both paraphrased here because they live in that other document). So a US operator gets no published window and an Australian one gets 60 days. Falsified by a US-side retrieval clause, most likely in the unpublished Master Subscription Agreement that the hardware leasing addendum names. https://gotab.com/australia-merchant-terms-and-conditions · retrieved 2026-09-05 adversarially verified

B
Yes

Published terms explicitly state the merchant owns its transaction and customer data, and constrain the vendor's right to use or resell that data (beyond aggregated/de-identified use) for its own products or marketing. differentiator

commercial-data-ownership-clause

Merchant retains ownership of Merchant Content; GoTab's use is confined to an aggregated and anonymized form for product improvement, analytics/benchmarking and Usage Data generation. GoTab separately markets itself as an Open Data platform where your guest data is yours to use. https://gotab.com/merchant-terms-and-conditions · retrieved 2026-08-01

B
Partial

Vendor documents validated P2PE or hardware-encrypted card entry plus tokenization such that the merchant's PCI scope is reduced to SAQ A / SAQ P2PE-HW, and names the applicable SAQ.

commercial-pci-p2pe-tokenization

Tokenization is documented (only a secure, encrypted token is stored) and card entry occurs on Adyen terminals, but no validated P2PE listing is cited and no merchant SAQ type is named. https://gotab.com/security · retrieved 2026-08-01

B
No

Vendor documents PCI DSS v4.0.1 future-dated controls in effect since 31 Mar 2025 — notably MFA for all access to the cardholder data environment including POS and back-office logins, and (for hosted ordering pages) payment-page script integrity monitoring per Req 6.4.3/11.6.1.

commercial-pci-dss-4-controls

Neither future-dated control is documented, and the access model that is documented is single-factor. This Getting Started article is the authority on GoTab logins: POS access is a PIN - the Server permission is "Required to pin into the POS and allows a staff member to take orders." - and manager dashboard access is a name-plus-phone check, managers reaching gotab.com/manager and "verifying with their name and cell phone"; the article's only security assurance is that PIN "numbers are one-way encrypted". Corpus-wide across gotab.com, success.gotab.io, gotab.readme.io and opsi.io, '\bSSO\b|\bSAML\b|single sign-on|multi.factor' returns 0 documents, 'MFA|two-factor|2FA' returns only opsi.io's own privacy policy (a separately-priced SKU, out of subject), 'cardholder data' returns 0, and 'script integrity|6.4.3|11.6.1|content security policy' returns 0. The only OTP endpoints in the public API are guest-facing (challengecustomer, verifycustomer). Falsified by an MFA enforcement setting or any Req 6.4.3/11.6.1 statement. https://success.gotab.io/knowledge/adding-users-and-creating-a-pin · retrieved 2026-09-05 adversarially verified

B
No

Product provides in-app tooling for CCPA/CPRA and GDPR data-subject requests — locating, exporting, and deleting an individual guest's records — plus a published DPA the operator can execute.

commercial-privacy-dsar-tooling

Both conjuncts fail. The privacy policy is guest-facing, framed as "part of the End User License Agreement", and its entire rights section is guest self-service - "You may delete your GoTab account at any time by visiting your customer profile page on the Web App and clicking the Delete Account link" - which is the guest acting on GoTab, not operator tooling to locate, export and delete an individual guest's records. There is no CCPA/CPRA section, no GDPR section, no controller/processor language and no DPA: corpus-wide 'data subject|DSAR|CCPA|CPRA|GDPR|do not sell|data processing (addendum|agreement)|subprocessor' returns exactly one document, opsi.io's own policy, which is a separately-priced SKU's legal document and cannot answer a GoTab contract cell. gotab.com/dpa, /gdpr, /ccpa and /subprocessors all 404 at the 66,059-byte control size, and merchant T&C 5.3 pushes all information-security and privacy-law compliance onto the merchant. The only deletion endpoints in the API remove a customer from a segment. Falsified by a published DPA or a guest-record erasure tool in the manager dashboard. https://gotab.com/privacy-policy · retrieved 2026-09-05 adversarially verified

B
Yes

POS supports card-network-compliant dual pricing / cash discount or surcharging with automatic exclusion of debit and prepaid cards from any surcharge, and required receipt and menu-board disclosure of the fee. differentiator

commercial-dual-pricing-compliant

The surcharging branch of this claim is fully documented: automatic exclusion — 'our surcharging feature does not apply surcharges to debit, prepaid, or gift card transactions'; network-cap compliance — maximum 3.0% (Colorado 2.0%) with CT/ME/MA/OK merchants ineligible; and disclosure — GoTab requires 'clear and visible disclosure signage at every entrance and point of transaction' and itself surfaces the fee 'on guest checks, receipts and Customer Facing Displays', with an itemized surcharge line on the printed receipt shown in the doc. GoTab also polices circumvention ('may disable any fees or service charges... intended to circumvent proper credit card surcharging'). Physical menu-board and entrance signage is executed by the merchant per the stated requirement. https://success.gotab.io/knowledge/credit-surcharging · retrieved 2026-08-05

B

Adversarial verification

An independent pass was instructed to refute this record, defaulting to downgrade when uncertain. It challenged 81 values — 62 upheld, 15 downgraded, 4 upgraded. It separately resolved 89 previously unknown values. This is published in full because a reader who can see which values were contested, on what evidence, and which way they moved has something no affiliate-funded comparison offers.

Pricing and identity

FieldVerdictWhat the verifier found
pricing.softwaredowngrade-to-partialThree errors against the live pricing page. (1) Loyalty + Memberships is +$35/mo on Pro AND +$35/mo on Sync — it is not included in Sync. (2) OPServer is 'Not available' on Basic, +$75/mo on Pro, included on Sync; the dossier omits that Basic cannot buy offline mode at all, which materially changes the offline story since the entire OPServer-dependent reliability cluster is unavailable at the headline $15 tier. (3) The opsi block includes an +$85/mo item alongside the +$45/mo modules, not uniform +$45. Fraud Protection, Easy Tab, 24x7 support, accounting and the rest check out as reported. source
pricing.hardwaredowngrade-to-unknownWrong URL checked. store.gotab.com is the storefront and carries individual hardware product pages. Whether prices are shown there is unverified — but 'not published per SKU' is asserted on a broken premise and must be withdrawn. source

Capability claims

ClaimAs first scoredVerdictWhat the verifier found
reliability-offline-card-authyesdowngrade-to-partialThree problems. (1) The sole sources are a marketing product page and the FAQ, not documentation. (2) The capability is gated: OPServer is listed on the pricing page as 'Not available' on Basic, +$75/mo on Pro, included only on Sync $229 — this is a paid add-on on a mid tier, not a platform property, and the dossier's own pricing block omits the Basic exclusion. (3) The mechanism is 'GoTab's standalone payment terminals continue processing credit and debit card payments offline' — 'standalone' is the Adyen term of art for a terminal operating detached from the POS, and Adyen's own docs describe store-and-forward for chip/contactless as a terminal-level feature. That is the acquirer's terminal doing S&F, not GoTab issuing an offline authorization, and it implies a degraded non-integrated flow. No per-transaction cap, no cumulative cap, no decline liability statement anywhere. This is on the highest-bar list and does not clear it. source
payments-offline-store-and-forwardyesdowngrade-to-partialSame single sentence, same defects as reliability-offline-card-auth: vendor marketing page only, gated behind the OPServer add-on that Basic cannot buy, mechanism attributable to Adyen terminal store-and-forward rather than GoTab, and the dossier itself concedes no caps are published. Cannot stand as 'yes'. source
kitchen-station-routingyesdowngrade-to-partialI fetched the KDS product page in full. It documents Smart Prep Times, color-coding by fulfillment time, ticket timers, bump bar and guest messaging — and never mentions station or zone routing. The phrase 'zone-based routing' appears only in the string 'Multi-state item tracking, smart batching, zone-based routing' on gotab.com/toast-vs-gotab, which is a vendor-authored competitive sales asset, the weakest evidence class in this exercise. The dossier's own note already conceded operator configurability is undocumented. A yes resting solely on a Toast-comparison page is not a yes. source
kitchen-sla-alertsyesdowngrade-to-partialThe KDS page supports only 'Orders are color-coded based on the time they should be fulfilled' and 'View ticket timers with all products and their respective prep times.' That is visual aging, not an SLA alert: no configurable threshold, no escalation, no notification, no breach reporting. The dossier's own note ('configurable per-station targets not detailed') describes a partial, not a yes. source
extensibility-free-sandboxyesdowngrade-to-partialThe Quick Start guide states: 'If you are not already a GoTab API User or do not have a sandbox account please contact api.support@gotab.io', and separately 'If your application needs to test credit card authorizations notify api.support@gotab.io'. Sandbox access is gated behind a support email, not self-serve, and nothing states it is free. This also falsifies the dossier's note on reporting-public-api ('self-serve credential creation and a sandbox') and its structural note claiming the docs are 'self-serve (no NDA, no sales call)' — the docs are self-serve to read, the sandbox is not self-serve to obtain. source
hardware-pricing-transparencynodowngrade-to-unknownThe affirmative 'no' rests on gotab.com/shop returning 404 — the wrong host. GoTab operates a separate storefront at store.gotab.com with per-SKU product pages that are search-indexed, including 'Mini Payment Terminal' (Adyen AMS1) and 'Full Size Payment Terminal' (S1F2). I could not load store.gotab.com or its product pages myself (404 to my fetcher, likely storefront/bot handling), so I cannot confirm whether prices are displayed — but the researcher's stated basis for the negative is invalid, so the cell cannot be scored 'no'. Same correction applies to pricing.hardware_cost. source
order-capture-floor-plan-editorunknownresolve-to-yesThe researcher never consulted GoTab's actual product documentation at success.gotab.io. It carries 'How do I Create a Floor Map?' and 'What are Maps?', documenting that 'Maps only allows you to map out a floor plan of the spots created in zones. Once a map is created, it will now show on the POS as an additional option to find the spots when initiating a tab', plus a Map View POS default view and 'Creating a Dine In Zone'. A drag-and-place floor-plan editor over zone-defined spots is documented. source
order-capture-void-comp-controlsunknownresolve-to-yesDocumented, not just inferred from the User Roles API. The comp/void article instructs 'Press comp or void at the top of the tab' then 'Select a reason' — reason codes exist. The Service (Employee Reports) page confirms manager gating and attribution: comps, voids and discounts each 'list out the employee that initiated' the action plus 'the manager that PIN'd in to approve'. Reason codes, manager approval and an exception report are all present. source
reporting-comps-voids-auditunknownresolve-to-yesService (Employee Reports) breaks out Comps, Voids and Discounts separately, each with initiating employee, item count, percent of sales, sales value, and the approving manager's PIN identity. That is a comps/voids audit report. source
labor-manager-override-auditunknownresolve-to-yesManager overrides are PIN-based and the approving manager is recorded and surfaced in reporting for comps, voids and discounts — 'the manager that PIN'd in to approve'. Override plus attributable audit trail is documented. source
reporting-server-scorecardspartialupgrade-to-yesThe dossier scored this partial-claimed off a vague 'monitor employee performance' line on the POS marketing page. The actual metric set is enumerated in product docs: per-employee sales with item quantity, guest count, percent of sales and total sales, plus comp/void/discount counts and percentages by employee, with server reports also listed under 'Server Necessities'. source
delivery-zones-polygonunknownresolve-to-yesThis is a flagged highest-bar cell and the researcher missed real evidence. Product docs cover 'Creating A Delivery Zone' and 'How To Create A Delivery Map': the operator draws the region in Google My Maps as a closed shape ('Type needs to be a single Polygon'), exports a .KML and uploads it to GoTab's Zones Dashboard. Genuine polygon delivery zones — though note the workflow is an external-tool KML import, not an in-product map editor, which is worth carrying in the note. source
delivery-zone-pricingunknownresolve-to-yesThe delivery zone configuration exposes per-zone commercial rules: a 'Minimum amount of purchases (in dollars) required to place an order' is set on the zone, and a delivery fee is configured alongside it. Caveat for the note: the minimum is explicit in the doc, the per-zone fee is referenced rather than fully specified, so if the rubric is strict this is a strong partial rather than a clean yes. source
order-capture-transfer-auditunknownupheldVerdict technically upheld only because the enum lacks upgrade-to-partial — this should move to partial. Tab transfer is documented ('How do I transfer a tab?', 'You can easily transfer any tab on the POS to another staff member', and tab reassignment to another server), so the capability exists. What remains genuinely undocumented is the two-employee audit logging the cell requires. source
order-capture-coursing-hold-firepartialupheldScore stands but the stated basis is wrong and should be corrected: the note claims 'named course objects and fire-next-course not documented', yet product docs carry a dedicated 'Full Service: POS Coursing' article under POS Ordering, alongside 'rushing orders'. Coursing is a named, documented feature rather than an inference from a handheld rush/delay control; I did not read the article deeply enough to confirm an explicit hold-then-fire action, so partial holds. source
reporting-public-apiyesupheldUpheld on the score — gotab.readme.io is genuinely public with no login, and the endpoint inventory the dossier cites is accurate: I confirmed 'Adjust or transfer inventory', 'Create/Update/Archive labor schedule', 'Bulk create products', 'Bulk update products', the full Wallet set, 'Trigger OTP Challenge'/'Verify OTP Code' and 'Refund a tab payment'. The note is wrong on one point: credential creation is not self-serve (see extensibility-free-sandbox). source
extensibility-webhooks-pushunknownupheldIndependently confirmed: the published documentation index contains only Guides, API Reference, Recipes and Changelog, with no webhooks, event-subscription or rate-limit section. Correctly held at unknown rather than 'no', since an index omission is not an affirmative vendor denial. Same for reporting-webhooks and extensibility-published-rate-limits. source
reliability-public-status-pageyesupheldVerified myself and the dossier understated nothing: login-free, per-component current status plus rolling 90-day uptime (Infrastructure 99.99%, POS 99.98%, KDS 99.98%, CFD 99.99%, QR Ordering/Kiosk/Print/Manager Dashboard/SMS/RFID/Payments 100.0%), dated incident history including the 24 Jul 2026 POS-CFD connection issue and a live degraded Twilio SMS dependency, with third parties tracked separately. source
commercial-autorenew-terms-publishedyesupheldConfirmed in the T&C: initial period specified on the Merchant Order Form, automatic monthly renewal thereafter unless either party gives 14 days' written notice. Also confirmed the 90-day fee-modification clause, the 30-day cure-based termination, absence of any software ETF or liquidated damages, merchant retention of Merchant Content ownership, the Stripe/Paysafe/Adyen processor set, and the absence of any uptime SLA or service credits — the dossier's commercial reading is accurate throughout. source
menu-pricing-channel-price-booksyesupheldConfirmed verbatim: 'Edit dollar amount, percentage markup per category and item or store for each marketplace within seconds', with the GoTab catalog flowing to third-party marketplaces. Per-channel price books are real. Note the page names no marketplace by name, which supports the dossier's separate partial on 3P coverage. source
delivery-driver-rosterunknownupheldHeld correctly even after finding the delivery zone documentation. The delivery articles cover zone geometry and order minimums only — the docs explicitly do not address driver assignment, a dispatch board, or driver tracking. Native in-house delivery dispatch remains undocumented, so delivery-dispatch-board, delivery-route-map, delivery-driver-tracking, delivery-driver-comp and delivery-cash-reconcile all correctly stay unknown. source
menu-pricing-modifier-price-by-parent-sizeunknown, grade F placeholder: 'No public documentation located during the 2026-08-01 research pass.'resolve-to-noThe modifier-creation doc enumerates the complete field set with a single flat 'Price: This amount will be added to the item's base price'; option groups price each option as one fixed amount, and Variant Details pricing applies to product variants, not to modifiers. No matrix by parent item or size exists anywhere in the documented catalog tooling. source
menu-pricing-topping-quantity-tiersunknown, grade F placeholder: 'No public documentation located during the 2026-08-01 research pass.'resolve-to-noGoTab's documented mechanism for multiple amounts of a modifier is Option Quantities — a min/max on selection count, the add-it-N-times pattern the claim explicitly excludes. No tier object (light/regular/extra/double) and no per-tier price multiplier appears in any modifier doc. source
menu-pricing-dual-pricingunknown, grade F placeholder: 'No public documentation located during the 2026-08-01 research pass.'resolve-to-noThe help centre documents credit surcharging — a percentage fee (max 3.0%, 2.0% CO) added on top of the single posted price at payment, debit/prepaid/gift excluded — as the card-cost-recovery mechanism. No cash-discount or dual-pricing mode is documented anywhere, and GoTab's own payment-processing explainer discusses dual pricing and cash discounting as strategies without claiming them as product. source
kitchen-printer-fallbackunknown, grade F placeholder: 'No public documentation located during the 2026-08-01 research pass.'resolve-to-noThe vendor's own support article for a down printer prescribes manual power-cycling of printer, switch and GoTab box while jobs queue as 'pending print jobs' until the printer returns; the KDS settings doc enumerates Devices & Printers settings (Fulfillment Printer, Closeout Printer, Pair to TDS) with no backup or failover field; POS notifications include no printer-offline alert. source
reliability-printer-fallbackunknown, grade F placeholder: 'No public documentation located during the 2026-08-01 research pass.'resolve-to-noSame evidence base as kitchen-printer-fallback: the enumerated KDS Devices & Printers settings contain no backup-printer option, the documented POS notification set (product availability, refund requests, QR ratings, QR scans) contains no printer-offline alert, and the printer-down support flow is manual power-cycling with jobs held as pending print jobs. source
labor-break-compliance-by-stateunknown, grade F placeholder: 'No public documentation located during the 2026-08-01 research pass.'resolve-to-noThe labor management doc enumerates break configuration in full — operator-created break types with a paid/unpaid toggle — plus four labor policies (Declare Tips on Clock Out, Prevent Clock Out with Open Tabs, Automatic Clock Out, Require Clock In). None of the claim's three components (per-jurisdiction rules, break attestation, missed-break premium flags) exists in the documented set. source
inventory-unit-conversion-yieldsunknown, grade F placeholder: 'No public documentation located during the 2026-08-01 research pass.'resolve-to-partialCore inventory is count-based with no unit concept (inventory-management doc: stock/par/reserved/available, recount/received/loss/theft). The opsi food-costing module — GoTab-owned since March 2025, sold as a $45/mo add-on — advertises 'built-in unit conversions and yield percentages'. Partial: capability exists first-party but only in a separately priced add-on, and the feature page does not document the purchase/recipe/count unit triple. source
reporting-pmix-modifier-levelunknown, grade F placeholder: 'No public documentation located during the 2026-08-01 research pass.'resolve-to-partialThe Product Mix report documents gross/net quantity and gross/net sales with a drilldown hierarchy that includes options (GoTab's modifiers) as a drill level. Partial: documented filters are date range, placed-vs-scheduled and user — no daypart or revenue-center/zone filter is documented on the report. source
reporting-tip-tax-complianceunknown, grade F placeholder: 'No public documentation located during the 2026-08-01 research pass.'resolve-to-partialDeclared cash tips are captured via the clock-out modal and a Payroll Report totals hours/rate/pay per user; charged tips and auto-gratuity appear in the Transaction Report CSV. Partial: no declared-vs-charged per-employee report, tip pooling only via the 7shifts integration, and no jurisdictional tax-liability summary documented. source
order-capture-throttlingunknown / F - Checked success.gotab.io/knowledge/pos, /knowledge/delivery-zone, and gotab.comresolve-to-partialThe order-throttles doc, unread by the prior pass, documents per-slot Order and Subtotal Limits on takeout/delivery zones with automatic slot closure at capacity. Partial: slots close instead of a quote-time extension, and limits are order-count-based, not kitchen-load-based. source
order-capture-scheduled-ordersunknown / F - Checked success.gotab.io/knowledge/pos, delivery zone docs, and gotab.com/producresolve-to-yesScheduled orders are documented across zone config (lead time, advance horizon, slot intervals), the POS takeout flow (date/time selection), and the KDS (scheduled-order display windows, hide-until-time-to-prep, takeout tickets fire when ready to prep rather than on receipt). source
order-capture-cateringunknown / F - Checked success.gotab.io help center POS category, gotab.com/products/online-ordresolve-to-partialEvent Deposits — advance payment via a deposit product/processor, later applied to the event tab, invoiceable by email with a due date and payable online — cover deposits and balance-due; quotes and a separate event schedule remain undocumented (Tripleseat is the events path). source
menu-pricing-countdown-auto-86unknown / F - Checked success.gotab.io/knowledge/inventory-management, gotab.com online-orderiresolve-to-yesDocumented countdown behavior: tracked stock decrements with orders and the '86 Product' out-of-stock action disables at zero and auto-re-enables the following day, with a named settings toggle (Re-enable Unavailable Products Overnight). All three claim elements are present at grade B. source
payments-dual-pricingunknown / F - Already scored 'no' with grade B evidence (success.gotab.io/knowledge/credit-surresolve-to-noPositive evidence of absence: catalog docs carry one flat price field, the documented card-cost-recovery mechanism is the credit surcharge, and Cash Tax-Inclusive is a tax-folding display toggle, not stored dual prices. Consistent with the verified no on menu-pricing-dual-pricing. source
payments-surcharge-guardrailsunknown / F - Checked success.gotab.io/knowledge/credit-surcharging. Documented: surcharge fearesolve-to-yesThe full program doc covers all three guardrails: automatic debit/prepaid/gift exclusion, the 3.0% cap (2.0% Colorado, four states ineligible), and per-merchant opt-in with auto-surfaced disclosure on checks, receipts and CFD. The prior pass had read this article but stopped short of scoring it. source
payments-tip-adjustunknown / F - Checked success.gotab.io labor management, payment processing, and gotab.com payresolve-to-yesAll four elements now sourced: preauth-then-capture tab model, a 72-hour documented adjust window on closed tabs (reopen path for older tabs), the POS View Tabs With No Tip screen, and QR/CFD tip prompts with configurable suggestions. source
payments-house-accountsunknown / F - Checked success.gotab.io/knowledge/add-house-account. Documented: House accountsresolve-to-partialHouse accounts are documented as internal payment processors with user/display account types, starting balances and liability reporting; credit limits and statement generation are absent from every house-account article, so partial with those named shortfalls. source
payments-payout-timingunknown / F - Checked gotab.com/pricing, merchant T&C, payment pages, and success.gotab.io payresolve-to-partialThe FAQ publishes a daily deposit cadence ('one consolidated deposit per day') in the food-hall context and the Payouts dashboard documents per-fiscal-day batches; no instant/same-day option and no general schedule for standard locations, so partial. source
kitchen-all-day-countsunknown / F - Checked success.gotab.io/knowledge/kds-settings and gotab.com/products/kitchen-dresolve-to-partialThe Multi-Items/All-Day panel is documented KDS functionality aggregating identical items across all on-screen orders; the doc itself states the panel omits notes and modifiers, hence partial. source
kitchen-recall-refireunknown / F - Checked success.gotab.io/knowledge/kds-settings, gotab.com KDS product page. KDSresolve-to-partialReset Order (unmark prepared/sent without re-ringing), Print To (reprint chit) and day-wide search across fulfilled tickets are documented; an unbump action returning a fulfilled ticket to the queue is not, hence partial. source
kitchen-speed-of-service-reportingunknown / F - Checked success.gotab.io KDS and reporting docs, gotab.com/products/manager-dashresolve-to-partialThe KDS Fulfillment Report (average processing time per day/time range plus top-10 slowest orders) was missed by the prior pass; no percentiles, station/channel slicing or export, hence partial. source
delivery-86-syncunknown / F - Checked success.gotab.io inventory, delivery, and POS docs, and gotab.com/producresolve-to-partialThe 3rd Party Ordering rejection FAQ confirms 86 status resyncs to DSP platforms and simultaneously documents the lag (orders rejected while 'resyncing is not yet complete'), which caps this at partial. source
delivery-store-pauseunknown / F - Checked success.gotab.io/knowledge/delivery-zone and online-ordering docs, and gresolve-to-noVendor-documented workaround is positive evidence of absence: the FAQ explicitly instructs operators NOT to use the in-GoTab zone toggle and to pause menus in the external aggregator platform instead. source
delivery-3p-reconciliationunknown / F - Checked success.gotab.io third-party ordering docs and gotab.com/products/onlineresolve-to-noVendor states payments and payouts for 3rd-party ordering are handled 100% independently of GoTab; without settlement data in-platform no payout reconciliation is possible and none is documented. source
digital-scheduled-pacingunknown / F - Checked gotab.com online-ordering, kiosk, and delivery pages, and success.gotab.resolve-to-yesZone-level scheduled ordering plus Throttles (order and subtotal caps per time step with automatic slot unavailability at capacity) match the claim; documented in the zones and order-throttles articles. source
digital-catering-portalunknown / F - Checked success.gotab.io help center and gotab.com products. Navigation referencresolve-to-partialCatering is assembled from documented generic parts (zone minimums/lead times, event deposits with online-payable invoices, Pack & Route for catered orders) with no dedicated portal, quotes or ACH terms, hence partial. source
digital-surcharge-transparencyunknown / F - Checked gotab.com online-ordering and kiosk product pages, and success.gotab.io resolve-to-partialThe surcharge program is platform-wide with auto-surfaced disclosure on guest checks, receipts and CFD plus jurisdiction and card-type handling; the QR checkout screen itself is not explicitly shown, hence partial. source
guest-loyalty-rfm-segmentationunknown / F - Checked success.gotab.io/knowledge/cart-rules-segments-loyalty-memberships and gresolve-to-noSegments are documented as manually built enrollment groups and the vendor's documented behavioral-segmentation path is Klaviyo; no auto-computed RFM/lifecycle segments exist in the product docs — positive evidence for no. source
guest-loyalty-lifecycle-automationunknown / F - Checked success.gotab.io loyalty docs and gotab.com customer engagement pages. Cresolve-to-partialTriggered lifecycle campaigns are delivered by the documented Klaviyo integration (real-time guest event streaming, flows built in Klaviyo), not natively — partial with the integration named as the shortfall. source
guest-loyalty-consent-managementunknown / F - Checked success.gotab.io loyalty and privacy docs, and gotab.com/privacy-policy.resolve-to-partialThe SMS opt-in doc covers checkbox consent capture with customizable disclosure, opt-out visibility, and STOP/START plus receipt-page revocation and re-consent; timestamp/source records and marketing-channel consent are undocumented, hence partial. source
guest-loyalty-campaign-attributionunknown / F - Checked success.gotab.io customer engagement docs and gotab.readme.io API refereresolve-to-partialThe Sales Page discounts drill-down ties every redemption to tabs via downloadable CSV with per-discount breakouts; incremental-lift measurement and send-level campaign attribution are absent, hence partial. source
guest-loyalty-review-capture-routingunknown / F - Checked success.gotab.io loyalty and customer engagement docs. No documentation resolve-to-partialA Customer Feedback report page, scheduled/placed feedback filtering and QR-rating POS notifications document capture and review; score-based routing to private recovery versus public review sites is undocumented, hence partial. source
labor-demand-labor-forecastunknown / F - Checked success.gotab.io/knowledge/howtomanagelabor and gotab.com manager-dashboresolve-to-partialNative output is a projected-sales figure in the Daily Insight email; demand-driven staffing lives in 7shifts, fed by GoTab's documented per-check sales export — partial via the named integration. source
labor-overtime-preventionunknown / F - Checked success.gotab.io labor docs and gotab.com products. Labor hours and poliresolve-to-noThe labor doc enumerates the complete four-policy set with no overtime threshold, and the only documented punch gate (7shifts clock-in enforcement) checks clocked-in status only — same enumeration basis as the verified break-compliance no. source
labor-tip-distribution-audit-trailunknown / F - Checked success.gotab.io/knowledge/howtomanagelabor and labor reports. 'Declare resolve-to-partialPer-employee tips received are natively reported (user reports, transaction CSV); the pool contribution/distribution ledger exists only in 7shifts via the documented check-level export, hence partial. source
labor-shift-swap-workflowunknown / F - Checked success.gotab.io/knowledge/howtomanagelabor and labor integration partneresolve-to-partialGoTab has no employee self-service app; scheduling and swaps are delegated to documented partner integrations (7shifts sync/enforcement/export, Schedulefly, Workforce.com), matching this record's partial-via-integration pattern. source
inventory-theoretical-vs-actualunknown / F - Checked success.gotab.io/knowledge/inventory-management and gotab.com products. resolve-to-partialopsi (GoTab-owned add-on) advertises exportable category-level variance reports with recipe counts and invoice-synced costs; category-level rather than per-item T-v-A and the add-on gating cap it at partial, grade C. source
inventory-86-auto-syncunknown / F - Checked success.gotab.io inventory docs and gotab.com online-ordering page. Manuresolve-to-partialAuto-86 at zero stock with overnight auto-restore is documented and the 86 state resyncs to DSP menus via the aggregator; trigger is product-level stock (no ingredient/recipe threshold in core) and 3P sync is laggy, hence partial. source
inventory-price-change-alertsunknown / F - Checked success.gotab.io/knowledge/inventory-management and gotab.com inventory resolve-to-partialopsi's food-costing page documents ingredient cost history, invoice-synced pricing and custom price-threshold notifications; add-on gating and feature-page evidence cap it at partial, grade C. source
inventory-par-auto-suggestunknown / F - Checked success.gotab.io inventory and manager dashboard docs. Par levels are maresolve-to-partialPar levels are documented but their entire function is a red low-stock flag; no suggested-PO generation or forecast mode exists in core GoTab or the opsi module lineup, hence partial on the par half only. source
inventory-shelf-life-expiryunknown / F - Checked success.gotab.io/knowledge/inventory-management. Stock tracking covers rresolve-to-noThe inventory doc enumerates the complete stock data model (recount/received/loss/theft, par, out-of-stock action, untrack, reserved/available) with no expiry or use-by concept anywhere in the KB or opsi modules — positive evidence for no. source
inventory-menu-margin-linkageunknown / F - Checked success.gotab.io inventory docs and gotab.com products. Menu items and iresolve-to-partialopsi's food-costing page documents margin/cost-summary reports and historical recipe profitability with invoice-synced costs; add-on gating, no margin-threshold flagging and no explicit sales-mix join cap it at partial, grade C. source
reporting-eod-closeoutunknown / F - Checked success.gotab.io/knowledge/manager-dashboard report list and gotab.com presolve-to-partialAll the closeout components are documented (sales drill-down and transaction CSV, cash counted-vs-expected reconciliation, payouts view, daily insight email, KDS closeout print) but no single reconciling EOD document, hence partial. source
reporting-cash-over-shortunknown / F - Checked success.gotab.io manager dashboard and payment reports, and gotab.com. Cresolve-to-yesCash reconciliation records counted-vs-expected per account with over/under flowing to user reports, cash accounting and the sales page; per-employee accounts and per-shift reconciliation are documented, satisfying the claim. source
reporting-channel-profitabilityunknown / F - Checked success.gotab.io/knowledge/manager-dashboard report list. Product mix, sresolve-to-partialZone-tagged revenue (transaction CSV Zone column, zones as channels, revenue_center mapping) is documented, but no per-channel margin report exists and 3P commissions never enter GoTab, hence partial. source
reporting-scheduled-deliveryunknown / F - Checked success.gotab.io manager dashboard and delivery docs. Delivery zone seturesolve-to-partialThe Daily Insight email is a documented scheduled report with recipient management and customizable content, but it is the only one and only daily — arbitrary reports cannot be scheduled, hence partial. source
reliability-cellular-backupunknown / F - Checked gotab.com/products/pocket-phone-only-pos (mentions optional SIM card), /resolve-to-partialThe handheld's optional SIM card is documented first-party cellular connectivity, but positioned for off-site events with no automatic-failover behavior and no cellular path for fixed terminals, hence partial at grade C. source
commercial-dual-pricing-compliantunknown / F - Checked gotab.com/pricing (card surcharge model documented), /merchant-terms-andresolve-to-yesThe claim's surcharging branch is fully documented: automatic debit/prepaid/gift exclusion, cap and state eligibility enforcement, and auto-surfaced disclosure on checks, receipts and CFD with itemized receipt lines; signage execution is the merchant's per the stated requirement. source
guest-loyalty-rfm-segmentationno / B - 'Segments - GoTab's only customer-grouping feature - are documented as manually b'upgrade-to-partialThe enumeration fails the bar: success.gotab.io/knowledge/segments is a how-to for creating one feature and never claims to enumerate GoTab's customer-grouping surface, so it cannot support 'only'. Counter-evidence found: GoTab acquired Fishbowl on 2026-06-09 specifically to 'segment audiences by frequency, spend, preferences, and loyalty status', and Fishbowl's own GRM page advertises Smart Segmentation over 'purchase patterns, dining frequency, and loyalty tiers' with machine-learning segment suggestion. Scored partial rather than yes because Fishbowl is a separate product, the deeper integration is described as forthcoming, and GoTab keeps it optional - the same treatment this record already gives opsi-delivered inventory capabilities. source
labor-overtime-preventionno / B - 'The labor doc enumerates GoTab's complete policy set - Declare Tips on Clock Out'upgrade-to-partialRe-read success.gotab.io/knowledge/howtomanagelabor: it documents at least eight labor settings, not the four the note calls the complete policy set (it also carries Print on Clock Out, Prevent Printing of User Report with Open Tabs, break-type paid/unpaid configuration and labor edit reasons), and it makes no completeness assertion - so the enumeration premise fails outright. Counter-evidence: 7punches clock-in runs inside the GoTab POS, and 7shifts' Overtime Warnings/Alerts require 'a POS labor integration or 7punches enabled' and notify when staff 'enter daily overtime, weekly overtime, or are at risk of weekly overtime', with a buffer that flags employees before they cross the weekly maximum. Partial with the shortfall named: the alert is manager-facing and does not warn or block the employee at the GoTab clock-in prompt. source
inventory-shelf-life-expiryno / B - 'The stock-level doc enumerates the complete inventory data model - Recount/Receiv'downgrade-to-unknownThe cited page does not meet the enumeration bar. It is a task-oriented walkthrough whose lists are introduced with 'you can do the following:' and 'can view and download the following reports' - available actions in the Manage Stock dialog, not a declared data model, and no completeness language appears on it. The note's remaining support is a nil result from sweeping 278 KB articles, which the evidence rules classify as unknown. I additionally checked opsi.io/products/inventory (the GoTab-owned add-on) for expiry, use-by, shelf-life, lot and FIFO and found none - consistent with the capability being absent, but still no positive statement of non-support anywhere, so 'no' is not supportable as written. source
labor-break-compliance-by-stateno / B - 'The labor doc enumerates the entire break feature: operator-created break types with a p'upgrade-to-partialThe enumeration premise fails: success.gotab.io/knowledge/howtomanagelabor documents six labor policies and break-type configuration, not the four the note treats as the documented set, and it never asserts completeness - so it can establish only absence-of-mention, which is not evidence of absence. Replacement evidence points the other way. 7shifts' GoTab POS integration article says open punches are sent on clock-in and 'Breaks, declared tips, and punch-out times are sent to 7shifts upon punch-out', satisfying the 'POS labor integration' precondition 7shifts states for its Labor Exceptions report; 7shifts' California Labor Compliance article documents a Jurisdiction setting (California, Berkeley, Emeryville), missed meal-break exceptions keyed to the 5th hour, missed rest-break exceptions, one hour of premium pay per exception, signed meal-break waivers and Employee Attestation - and its help center carries the same for New York, Washington, Illinois, Oregon, Nevada, D.C. and Philadelphia. Scored partial rather than yes because the entire compliance layer is a separately contracted 7shifts plan (GoTab's setup doc requires 'Entree Plan or higher', 7shifts warns of a further upgrade), attestation happens in the 7shifts punch modal rather than at the GoTab clock-out prompt, and exceptions are reported rather than enforced at the terminal. source
labor-qualified-tips-w2-reportingunknown, grade F: 'no doc mentions qualified tips, W-2 Box 12 code TP, Box 14b, or Treasury tipped-occupation codes'resolve-to-partialThe refuted premise cut the other way here: because howtomanagelabor is not exhaustive I went to the partner side and found the documentation the earlier pass missed. 7shifts' GoTab POS article enumerates three tip sources GoTab supplies - CC Tips, Auto-gratuity and Declared Tips - and states they feed 7shifts Payroll and external payroll exports, which is the cash-versus-charged qualified-tip separation the claim asks for; the earlier rationale's 'tips only as type: total = CC Tips' reading came from the tip-pooling receipt payload alone and missed the punch-borne declared tips. The W-2 half fails: the SOC code is hand-entered in 7shifts Payroll for state unemployment returns, and the OBBBA article covers only tax year 2025 via Box 14 Qualified OT and a Box 7 tip estimate, with nothing on Box 12 code TP or Box 14b. Partial, shortfall named: separation yes, coding no, and all of it downstream of GoTab in a separately purchased product. source
labor-tip-distribution-audit-trailpartial, grade B - 'Per-employee tip receipts are recorded natively - POS user reports comput'upheldValue unchanged; the note understated both sides. It said only CC tips ride the 7shifts export, but the 7shifts GoTab POS article states breaks, declared tips and punch-out times are all sent on punch-out and names three supported tip sources (CC tips, auto-gratuity, declared tips) usable in Tip Pooling with distribution by hours, points or percentages, flowing into 7shifts Payroll and external exports. It also missed GoTab's own Labor page reports, including a Download Audit - which is a punch-edit audit, not a pool audit, so the shortfall survives. Re-stated with the plan and configuration gates named. source
labor-payroll-export-formatspartial, grade B - 'Labor integrations named are Workforce.com, 7shifts, Schedulefly, TipHa'upheldValue unchanged, reasoning replaced. The note rested on an integration-directory enumeration - the shape this audit disqualifies - and it missed GoTab's own Labor page, which publishes a Download Payroll Report of total hours, rate and pay per user plus a punch report and a punch-edit audit. The shortfall is different from the one recorded: not the absence of named payroll partners in a directory, but that no file format is documented, no payroll-provider connector exists (QuickBooks Online is GL, not payroll), and payroll delivery runs through 7shifts. source
order-capture-native-handheldyes, grade B - First-party handheld POS app on Samsung A17; takes orders, rupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
order-capture-offline-order-entryyes, grade B - OPServer keeps POS, handhelds and kiosks creating orders andupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
order-capture-kiosk-first-partypartial, grade B - First-party kiosk shares the product catalog with no separatupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
menu-pricing-channel-price-booksyes, grade B - Confirmed verbatim: 'Edit dollar amount, percentage markup pupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
menu-pricing-3p-menu-pushpartial, grade B - GoTab catalog and menu items flow to third-party marketplaceupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
payments-pay-at-tableyes, grade B - Handheld takes orders and payments tableside; Easy Tab dips/upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
payments-offline-store-and-forwardpartial, grade B - Same single sentence, same defects as reliability-offline-caupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
payments-gift-cardsyes, grade B - Native gift cards as a $35/mo add-on; customer engagement paupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
kitchen-station-routingpartial, grade B - I fetched the KDS product page in full. It documents Smart PupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
kitchen-sla-alertspartial, grade B - The KDS page supports only 'Orders are color-coded based on upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
delivery-menu-pushyes, grade B - Catalog and menu items flow from GoTab to third-party marketupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
delivery-tracking-pagepartial, grade B - Guests get live order updates on their tab and two-way SMS; upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
digital-first-party-webyes, grade B - Commission-free ordering accepted from your website, social upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
digital-menu-single-sourceyes, grade B - One product catalog drives kiosk, QR, POS and online — no seupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
digital-qr-tableyes, grade B - Scan-to-order and scan-to-pay attach to the same open tab, wupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
digital-group-orderingpartial, grade B - Multiple guests join and order on one shared tab and split tupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
digital-loyalty-attachyes, grade B - Members are automatically recognized, no separate cards to supheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
guest-loyalty-accrual-modelsyes, grade B - Offers can key on specific item purchases, spend amounts or upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
guest-loyalty-native-email-smspartial, grade B - Two-way SMS is native; email marketing is described as synciupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
guest-loyalty-stored-value-giftyes, grade B - Native gift cards (+$35/mo) digital and physical, plus RFID upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
inventory-recipe-bom-costingpartial, grade B - Delivered by partners — opsi recipe/inventory modules at +$4upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
inventory-native-not-partnerno, grade B - GoTab has basic native stock tracking, but recipe costing anupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
reporting-realtime-dashboardyes, grade B - Manager Dashboard gives real-time and historical sales, acceupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
reporting-tier-paywallpartial, grade B - Basic at $15/mo excludes accounting integration (+$10), multupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
hardware-os-platformspartial, grade B - Windows, iOS and Android are named on the pricing and FAQ paupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
hardware-handheld-purpose-builtpartial, grade B - The handheld is a Samsung A17 consumer phone purpose-built fupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
hardware-offline-modeyes, grade B - OPServer keeps terminals ordering, printing and taking cash upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
hardware-kdsyes, grade B - First-party KDS sold at +$25-$35/mo with bump bar, smart preupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
hardware-selfpour-scalesyes, grade B - Documented self-pour and beverage-hardware integrations: iPoupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
reliability-offline-order-entryyes, grade B - Order entry, printer routing and KDS continue on-premises viupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
reliability-offline-card-authpartial, grade B - Three problems. (1) The sole sources are a marketing productupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
reliability-offline-kds-printingyes, grade B - KDS continues operating without interruption and orders keepupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
reliability-offline-feature-matrixpartial, grade B - Vendor affirmatively names QR ordering, third-party labor syupheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
reliability-247-live-supportpartial, grade B - 24x7x365 support via email, phone, chat and in-app exists — upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
commercial-interchange-plus-publishedno, grade B - Only a blended card-present rate is published (starting at 2upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
commercial-pricing-publishedyes, grade B - Specific dollars published: Basic $15/mo, Pro $99/mo, Sync $upheldGrade only, value untouched. Corpus-wide mechanical pass (workflows/regrade-marketing-pages.mjs): the cited page is a first-party pricing/feature/product page, which atlas-evidence-rules grades C ("vendor pricing page or feature page - first-party, but written to sell"), not B ("vendor product documentation - the manual, the help centre, the admin guide"). No claim was re-researched in this pass and no evidence was re-read; only the grade was brought into line with the source type. source
order-capture-bar-tab-preauthyes / grade B, cited to gotab.com/products/insufficient-funds-protection: 'Easy Tab opens a tab by dipping/tapping the card; Fraud Protection re-authorizes on each new order'upheldThe marketing product page was replaced with the help-centre documentation, which is more specific than the marketing wording and closes the gap the original note conceded. Insufficient Funds Protection gives the four Incremental Auth on Send Order settings and the fund-source options and their location in the manager dashboard; Location Settings: Edit gives the Pre-authorization Amount and Preauth Initial Order fields; End User Experience: How to Open a Tab states the automatic end-of-business-day close that the first pass recorded as undocumented. All three elements of the claim are now on grade-B pages. source
order-capture-qr-same-checkyes / grade B, cited to gotab.com/products/mobile-ordering-payment: 'Core architecture: staff and guests share the same tab, so guest QR items land on the same p'upheldThe original note paraphrased a marketing page's architecture pitch. The help-centre Easy Tab article says the same thing operationally and in more detail: the tab a guest orders on from their phone is the tab reachable from the POS, servers add orders to it, and the two sources of items are identified within that single tab. That is documentation of same-check behaviour, not a repeated sales line. source
payments-published-ratesyes / grade B, cited to gotab.com/pricing: 'Starting at 2.40% + $0.15 card-present; +0.85% Amex; +1.5% international. Published publ'downgrade-to-partialThe citation is the pricing page, which the grade table fixes at C, and a yes on a differentiator needs A or B. Searched the help centre (terms: processing rate, fees, surcharging, a la carte) and docs.gotab.io: neither publishes a processing rate, and the Credit Card Surcharging article covers surcharge configuration, not GoTab's own pricing. The Merchant Terms and Conditions define Fees by reference to the Merchant Order Form. What is genuinely published is a qualified starting rate plus two published uplifts, which is materially more than quote-only and materially less than a published schedule - partial with the shortfall named. source
payments-qr-guest-payyes / grade B, cited to gotab.com/products/mobile-ordering-payment: 'Guest scans QR, views the live tab, tips and checks out from their own phone, closing the s'upheldRe-cited off the marketing product page onto the end-user help-centre articles, which describe the steps rather than assert the outcome: scan at table, review tab, tip selection, payment-method selection, close tab, receipt. Same-check identity comes from the Easy Tab article. The one wrinkle is that GoTab's documented entry point is the table/spot QR (and a bill texted to the guest per the Starting a Dine In Order article) rather than a QR printed on a paper check; the capability the claim describes - guest views and pays their own check from their phone and it closes in the POS - is documented. source
kitchen-prep-time-pacingyes / grade B, cited to gotab.com/products/kitchen-display-system-kds: 'Smart Prep Times - longer-prep items display earlier so items reach the pass together; tick'downgrade-to-partial'Smart Prep Times' is a marketing name; the KDS documentation does not describe staggered item start or display times. KDS: Ticket Timer documents a single ticket-level countdown seeded by the longest item prep time. KDS: How to Sort Items on your Tickets documents By Prep Time ascending/descending as a sort order only. The genuine cook-time-aware staggering GoTab documents is between courses, via the Full Service: Coursing Configuration auto-timed formula, and it requires the coursing feature to be enabled by GoTab for the location. Real capability, narrower than the claim, so partial with the shortfall named rather than yes. source
kitchen-offline-operationyes / grade B, cited to gotab.com/products/on-premise-server-with-enhanced-offline-mode: 'KDS continues operating without interruption and orders keep routing to printers during an'downgrade-to-partialThe cited page is the OPServer product marketing page. Probed the documentation hosts: success.gotab.io site search returns 0 results for 'OPServer' and nothing relevant for 'on-premise'; docs.gotab.io has no offline or OPServer material. The strongest documentation is the GoTops app article, which establishes a local POS/KDS client usable in an outage but says nothing about ticket delivery, and the 5/9/25 release note, which explicitly defers sending the order until reconnection. Since the ordering path stalls at send in the documented base behaviour, the KDS cannot be shown to keep receiving tickets. Partial, shortfall named. source
reliability-lan-degraded-multi-terminalyes / grade B, cited to gotab.com/products/on-premise-server-with-enhanced-offline-mode: 'All POS terminals, handhelds, kiosks and payment devices remain connected through a shared'downgrade-to-unknownThe whole basis was a marketing product page, which the grade table puts at D-C and which cannot carry a yes on a differentiator. Re-probed both documentation hosts with the help centre's own search API (/_hcms/search) rather than relying on navigation: OPServer has no article, and no article anywhere describes local check-state sharing between terminals. Not downgraded to no, because the OPServer add-on plausibly does this and absence of evidence is unknown - but there is nothing to score a value on, so grade F with no url. source
reliability-local-transaction-engineyes / grade B, cited to gotab.com/products/on-premise-server-with-enhanced-offline-mode: 'OPServer is a named on-premise server product, +$75/mo and included in the Sync tier, that'downgrade-to-partialThe claim asks whether the vendor DOCUMENTS a local transaction engine. A product page naming and pricing one is a marketing claim, not documentation, and the search of both documentation hosts found no OPServer article at all. What documentation does establish is a locally installed POS/KDS application usable during an outage, which is a real but weaker architectural fact than an edge transaction engine. Partial with the gap named rather than yes on a sales page, and not unknown, because the GoTops client is documented. source
commercial-module-unbundlingyes / grade B, cited to gotab.com/pricing: 'The price sheet is explicitly a la carte - KDS, kiosk, loyalty, gift cards, aggregation, m'downgrade-to-partialThe evidence is the pricing page, which is grade C, and a differentiator yes needs A or B. Re-read the page: it is an add-on price grid layered on three plans, and the Basic plan marks five of the modules 'Not available', so the modules are not universally purchasable a la carte. The second half of the claim - cancel independently without repricing the base subscription - is a contract term, and the Merchant Terms and Conditions publish only whole-agreement term, monthly renewal and termination-for-breach language with fees deferred to the Order Form. Independent per-module cancellation is unpublished. Partial with both shortfalls named. source
order-capture-seat-levelunknown / F - Seat-number tagging not documented. Guests self-split a shared tab, a different mechanism."resolve-to-yesupheld source
order-capture-transfer-auditunknown / F - Verdict technically upheld only because the enum lacks upgrade-to-partial — this should move to partresolve-to-partialupheld source
order-capture-order-ready-signalunknown / F - Read the full 3rd Party Ordering setup doc (success.gotab.io/knowledge/3rd-party-ordering, 2026-08-0upheldoverturned: no -> unknown; a setup guide that never takes up outbound status is not an enumeration, and the aggregator portal that would carry the signal is outside the corpus source
payments-offline-decline-liabilityunknown / F - No public statement of who bears the loss on a stored offline transaction that declines, and no postresolve-to-noupheld source
payments-refund-void-controlsunknown / F - API exposes Refund a tab payment and User Roles, but role-gating of refunds/voids and an immutable aresolve-to-partialupheld source
payments-chargeback-toolingunknown / F - Insufficient Funds Protection is described as minimizing failed-payment chargebacks with reporting aresolve-to-nodowngraded: partial -> no; the vendor FAQ answers the cell in the negative and the only operator channel is email source
payments-p2pe-pci4unknown / F - Security page documents tokenization and end-to-end industry-standard encryption but names no PCI DSupheldupheld: url repaired from undefined to the page the reasoning rests on source
reliability-offline-decline-liabilityunknown / F - No public statement of decline liability and no per-transaction or cumulative offline cap is publishresolve-to-noupheld source
reliability-lan-degraded-multi-terminalunknown / F - Searched GoTab's help centre (success.gotab.io site search: 'OPServer' returns 0 articles, 'on-premiresolve-to-partialupheld: partial/C is consistent with the standing reliability-offline-card-auth and payments-offline-store-and-forward rows, which hold the same page to partial/C for the same add-on gating and marketing sourcing source
reliability-sync-conflict-handlingunknown / F - GoTab says data securely syncs back to the cloud automatically with no data loss, but publishes no cresolve-to-noupheld: a documentation-existence claim, unlike failover-terminal-role which is a capability claim - the OPServer fact sheet is form-gated and no directly-linked OPServer PDF exists on the partner resource centre source
reliability-failover-terminal-roleunknown / F - OPServer is the designated local master; whether a terminal can assume that role automatically if OPupheldupheld: url repaired from undefined to the page the reasoning rests on source
delivery-driver-rosterunknown / F - Held correctly even after finding the delivery zone documentation. The delivery articles cover zone resolve-to-noupheld source
delivery-address-validationunknown / F - Read the delivery articles in full (2026-08-05): the operator draws the delivery area in Google My Mupheldupheld source
delivery-injection-error-visibilityunknown / F - Read the full 3rd Party Ordering doc plus pos-notifications (2026-08-05). Injection failures demonstupheldoverturned: no -> unknown; the manager-app notification list is unenumerated and the aggregator portal is unread source
digital-apple-business-connectunknown / F - Checked the full 278-article KB sitemap (2026-08-05): the documented map-placement path is Google ('resolve-to-noupheld: re-checked over the clean 877-doc index - apple business connect, apple maps and place card still return zero source
guest-loyalty-thirdparty-identity-attachunknown / F - Checked the help centre's third-party ordering setup doc (success.gotab.io/knowledge/3rd-party-orderupheldupheld source
guest-loyalty-offline-behaviorunknown / F - The OPServer offline page names QR ordering and third-party labor as internet-dependent but says notresolve-to-noupheld: the FAQ eleven-question offline block is a genuine enumeration with the loyalty column empty; all quotes verified in gotab.com/faq source
guest-loyalty-10dlc-registrationunknown / F - Not documented. The status page tracks Twilio SMS Long Code North America as a dependency, implying resolve-to-yesupheld: the cell reads "handles or explicitly documents", and openphone-account-setup documents brand and campaign registration step by step - note the sibling sms-platform-transition records that GoTab discontinued outbound messaging on 1 Nov 2024 for operators who had not transitioned, so GoTab does not itself register on the operator behalf source
guest-loyalty-wallet-passunknown / F - No Apple/Google Wallet loyalty pass documented. RFID Pass is a physical wristband/card, a different resolve-to-noupheld: wallet pass, passkit and pkpass re-checked at zero over the clean 877-doc index source
guest-loyalty-privacy-rights-toolingunknown / F - Swept the full KB sitemap (278 articles, 2026-08-05): customer-data surfaces are the Subscribers pagupheldoverturned: no -> unknown, and re-homed off gotab.com/privacy-policy (the "right to delete" string was in www.opsi.io/terms, another company) source
guest-loyalty-ai-offer-recommendationunknown / F - Checked the 278-article KB sitemap, the customer-engagement and memberships product pages, the annouresolve-to-noupheld: a claim-level cell, and the absence of the claim from the vendor own engagement enumeration is close to dispositive; the machine-learning hits are correctly attributed to opsi.io in the note source
inventory-shelf-life-expiryunknown / F - Assessed and unresolved. GoTab's stock-level article (success.gotab.io/knowledge/inventory-managemenupheldoverturned: no -> unknown; a marketing product page with six feature blocks is not a data-model census and never takes up dates at all source
reporting-webhooksunknown / F - GoTab's complete published doc index contains no webhooks or event-subscription section; the documenresolve-to-partialupheld: push to a customer-supplied endpoint is real and carries payment data, but signature verification and retry both fail, so partial not yes source
reporting-history-retentionunknown / F - Checked (2026-08-05) the Manager Dashboard report docs (sales, PMIX, fulfillment, payments — all offresolve-to-noupheld: a documentation claim - no retention window, truncation floor or archive fee is stated anywhere on the four hosts source
reporting-anomaly-alertsunknown / F - Checked (2026-08-05) pos-notifications (product availability, refund requests, QR ratings, QR scans upheldoverturned: no -> unknown, and re-homed onto the document that actually carries the notification language; an unenumerated list forbids a no source
reporting-nl-queryunknown / F - Checked (2026-08-05) the full KB sitemap including every product-update note through 4/27/26, the Maresolve-to-noupheld: natural language, copilot and chatbot re-checked at zero over the clean 877-doc index source
reporting-guest-cohortsunknown / F - Checked (2026-08-05) the payments dashboard doc (per-payment guest detail with CSV export, no guest resolve-to-nore-homed: off getting-started-sales onto the KB category page that enumerates the whole Reports set - one of the 17 category pages added to the index after the analysts ran source
reporting-sales-forecastunknown / F - Not native. BI partners ProfitRover, Tenzo, DYNE, Avero and Five x 5 appear in the directory but GoTresolve-to-partialupheld: projected sales is a genuine native forward figure (the token projected occurs in only two documents corpus-wide) but daypart granularity, UI exposure and scheduling consumption all fail; ProfitRover is correctly barred from scoring the cell source
extensibility-webhooks-pushunknown / F - Independently confirmed: the published documentation index contains only Guides, API Reference, Reciresolve-to-partialupheld: P2 is dead and partial is the right landing - real push to a customer-supplied URL, but the only order-state event is the closed tab and the endpoint is partner-configured, not operator-subscribable source
extensibility-webhook-reliabilityunknown / F - Moot unless webhooks exist; no HMAC signing, retry-with-backoff or replayable event log is documenteresolve-to-noupheld: the loyalty API enumerates what the integrator must supply and the security item is a static Authorization header, not a signature; retry-after and idempotenc re-checked at zero over 877 source
extensibility-middleware-compatibilityunknown / F - None of Deliverect, Chowly, Otter, Checkmate or ItsaCheckmate appear in GoTab's integration directorupheldoverturned: no -> unknown; the surface that answers this cell is the middleware vendors own catalogues, which were never in scope source
hardware-handheld-battery-swapunknown / F - Checked (2026-08-05) the Pocket POS product page: the handheld is a Samsung A17 consumer phone markeresolve-to-noupheld: corroborated by the shop registry nobody had - 26 priced peripherals including A17-specific accessories but no A17 battery, and the one replacement battery SKU (S1F2BATT, tagged S1F2 Accessories) belongs to the Adyen payment terminal, not the handheld source
hardware-handheld-lteunknown / F - The Samsung A17 is a cellular phone, but GoTab does not document LTE as a supported connectivity falresolve-to-partialupheld: corroborated by the shop registry, which sells a T-Mobile sim card SKU with a monthly service charge - still off-site cellular, still no failover behaviour source
hardware-peripheralsunknown / F - A Customer Facing Display component appears on the status page, but no published compatibility list resolve-to-partialre-homed: the "optional accessories" quote was in gotab.com/faq, not the cited article; the verdict now rests on GoTab own product registry source
hardware-pricing-transparencyunknown / F - The affirmative 'no' rests on gotab.com/shop returning 404 — the wrong host. GoTab operates a separaresolve-to-yesoverturned: unknown -> yes on the registry nobody had; four hosts had been named for this cell and all four were wrong source
hardware-usable-after-churnunknown / F - Leased equipment remains the sole property of GoTab unless and until the Hardware Balance is paid inresolve-to-partialupheld: the missing MSA is a missing document, not a missing practice, and the note already says so; the leased path is settled by clause 5.1/5.2 and the purchased path is genuinely unaddressed source
reliability-onsite-installunknown / F - The limited warranty references installation completed by a GoTab representative as a condition, impresolve-to-partialupheld source
reliability-menu-build-serviceunknown / F - A setup fee customized to your unique needs and financeable implementation costs imply onboarding seresolve-to-noupheld: corroborated by the shop registry, whose entire Services category is two SKUs, QR Setup Fee and QR Code Stickers, with no menu-build line; the clover comparison page says only that onboarding handles setup and never pins that to menu entry source
reliability-pci-dss-4-attestationunknown / F - The security page documents tokenization and encryption but publishes no PCI DSS v4.0/4.0.1 AoC or Pupheldoverturned: no -> unknown; a directly-linked, un-gated PCI PDF that nobody fetched cannot ground a no about publication, and gotab.com/security is the wrong subject source
commercial-post-termination-export-windowunknown / F - The merchant T&C addresses post-termination confidentiality but specifies no data-retrieval window."resolve-to-partialupheld: a jurisdiction split is the accurate finding, not an artefact - clause 2.4 is a published GoTab contract term and the note already flags that the US terms carry no counterpart source
commercial-pci-dss-4-controlsunknown / F - No mention of PCI DSS v4.0.1 future-dated requirements — MFA for CDE access or payment-page script iresolve-to-noupheld: unlike the attestation cell, this one is answered by affirmative product evidence - the documented POS login is a four-digit PIN, which cannot satisfy MFA for CDE access whatever the unread PCI PDF says; SSO, SAML, multi-factor and cardholder data all re-checked at zero over 877 source
commercial-privacy-dsar-toolingunknown / F - A privacy policy exists at gotab.com/privacy-policy, but in-app CCPA/CPRA/GDPR data-subject request resolve-to-noupheld: unlike guest-loyalty-privacy-rights-tooling, this cell carries a second conjunct that is control-measurable - gotab.com/dpa returns the 66,059-byte 404 and no DPA exists anywhere in the corpus, so the claim fails on a published absence rather than an unenumerated one source
delivery-3p-direct-integrationpartial / B - Uber Eats, Deliveroo and Just Eat are named partners. DoorDash appears only in an FAQ aside and is aupheldnew: standing note falsified source
delivery-address-validationunknown, grade Fresolve-to-partialRe-opened every cited artefact myself. The bundle is authentic and live: I re-fetched https://cdn.gotab.io/static/uuid_1-yi~42832916.ea3dfe34.js today and its SHA256 (F80F357EBEEE5DA56F25A2490493D804F0B5F450A1E84AEFC90A9F2E61BB613A) is byte-identical to the local copy. All four quoted items confirmed verbatim: the {libraries:["places"]} loader wrapping the component, the nine-entry address_components map in onGoogleMapsReady, the string 'Please select an address by typing in the field above.' (present twice, gated on show_autocompleted_error), and the zones/days call appending address_uuid. The proposer UNDERSTATED the strongest item: addCustomerAddress does not merely POST when autocompleted, its else-branch sets show_autocompleted_error=true, and a keyup on the input clears autocompleted — a hard gate, not a soft preference. Also verified checkAddressFields(){...show_autocompleted_error=!street_1||!city||!zip}. The docs page cited for the KML map is a real route (nonsense sibling slug 404s, so not a soft-200) and carries 'Click upload a .kml file to create your delivery map in GoTab' and the single-Polygon/Linestring warning verbatim. Kept at partial rather than rejecting: the claim is a conjunction, and the first conjunct — geocoded/validated against a mapping service at order time — is met outright at grade A, which is also the conjunct the claim's own why_it_matters rests on ('an unvalidated address is a food-cost loss'). It is not self-refuting because it does not shortfall every conjunct. But I cut the proposal back in three ways. (1) DOUBLE COUNT: its KML/polygon half re-uses the proposition already banked at delivery-zones-polygon (yes/B, success.gotab.io/knowledge/delivery-zone, now mirrored at docs.gotab.io), and operator-side geometry cannot evidence address-side enforcement, so I removed it from the note. (2) I ran the negative controls myself over corpora the proposer did not search in the form they implied — their notes.json 'searches' would have hit only source DESCRIPTIONS, since notes.json values are one-line summaries, not source text; the real 952 KB llms-full.txt, the 326 KB OpenAPI spec and the 373 KB GraphQL schema all had to be fetched fresh. They confirm the shortfall and strengthen it: no coordinates on Address, no geometry on Zone. (3) TRAP NAMED: goGetAddressLocation(_addressId) returns a bare boolean and belongs to a family with goGetItemLocation and goGetInventoryAdjustmentLocation — it is Postgraphile ownership/RLS plumbing, not a geocoder or a zone test, and reads exactly like the evidence this cell lacks. Substring traps checked and held: tokenised /\bzone\b/ gives 105 genuine hits in the JS and 229 in the KB (timezone cannot match a leading word boundary, so the zone concept is real, not a timezone artefact), and the ad-markup 'delivery' the proposer flagged is confirmed unrelated IAB VAST attributes. Grade A stands: shipped source code, retrieved by me, and neither cdn.gotab.io/robots.txt nor docs.gotab.io/robots.txt exists (both 404), so no bar applies. CITATION CONFLICT, flagged deliberately: final_url is NOT a key in notes.json, because no key in notes.json carries this proposition — the only URL that does is the CDN bundle, which this session recorded in .cache/q310-gotab/js-index.json (115 entries) rather than notes.json. Citing a notes.json key here would have meant citing the GraphQL schema, which evidences the shortfall and not the capability. The evidence rule to record the URL I actually retrieved outranks the output constraint, so I cite the bundle and name the deviation. Prior value was unknown, so when this is applied to verification.jsonl it is a resolution (resolve-to-partial), not a downgrade. source
guest-loyalty-thirdparty-identity-attachunknown, grade FupheldProposed partial from the docs.gotab.io OpenAPI spec, GraphQL schema and guest ordering bundle (310th session); rejected on verification: REJECTED; the cell stays unknown/F on its standing verified rationale, which I re-read and found still accurate. When applied, the row keeps rationale and NO url, per the evidence rules for an unknown. EVERY QUOTE VERIFIED, AND THEY ALL HOLD. I re-fetched https://docs.gotab.io/specs/gotab-api.json (200, 326,626 bytes, OpenAPI 3.1.0, 'GoTab API' 2.0, 65 paths, 72 schemas) and https://docs.gotab.io/graphql-schema.txt (200, 373,184 bytes, 9,181 lines) today and parsed both. TabBody.phoneNumber is verbatim 'phone number to associate with the tab. Must be provided if `openTab` is false and a `customerId` is not provided.'; TabBody.customerId verbatim 'Optional customer id to associate with the tab. Must be provided if `openTab` is false and a `phoneNumber` is not provided.'; evaluateOrderRules verbatim. Tab.customerId (line 1031) and Tab.segmentIds (1032) exist; Order.guest: Guest (993); Guest (1691) is exactly {name, customerId, locationId, locationUuid, tabUuid, feedback, handle, email, userId, tabViewed} under the comment 'A resource representing a Guest on a Tab'; ExternalSourceEnum (8935) carries STREAM_DSP among 11 members; DbExternalId (2322) is {id, dbExternalId, xSource, locationId, tableName, dbExternalIdsPrimaryKey, xId, location}. So the proposal is not fabricated. It is misread. THE REJECTION IS ABOUT SUBJECT, NOT ACCURACY. The claim is conjunctive and marketplace-specific: marketplace orders 'attach to a native guest profile WITH USABLE GUEST IDENTITY rather than landing as anonymous tickets', and its why_it_matters is migrating marketplace guests to owned channels. (1) The proposer CONCEDES the identity conjunct outright -- 'nothing in either published surface shows the marketplace actually SUPPLIES usable guest identity'. (2) The surviving conjunct is not marketplace evidence at all. TabBody's phone-or-customerId requirement governs EVERY closed tab created through the API -- kiosk, reservation, integrator web checkout -- and is exactly the generic hygiene the brief names. I ran the binding control the proposer did not: across all 326 KB of the spec, DOORDASH = 0, UBER = 0, GRUBHUB = 0, marketplace|aggregator|third[- ]party|3rd party = 0, STREAM_DSP|externalSource|xSource = 0, API_Integration = 0. The spec never binds itself to the marketplace path; that binding is supplied entirely by the proposer's inference that Stream calls this API, and Stream is opaque third-party software whose payloads nobody here has observed. (3) evaluateOrderRules actively CUTS AGAINST the proposal: it defaults false and its own description ends 'Used by the kiosk loyalty/membership check-in flow' -- an in-venue affirmative check-in, which is precisely what a DSP-injected order cannot perform. The proposer quoted the sentence's first half and dropped the half that locates it. (4) DbExternalId/STREAM_DSP is ORIGIN TAGGING, NOT IDENTITY. Its shape is {xSource, tableName, xId} per location -- a generic row-mapping table keyed by database table name, shared with MEWS, XERO and 7SHIFTS. It proves a Stream-origin record is labelled; it says nothing about a guest. (5) Guest is per-tab by its own comment, as the proposer concedes, so it cannot be the 'native guest profile' the claim requires. THE STRONGEST EVIDENCE IS ONE THE PROPOSER MISSED, AND IT STILL FAILS. docs.gotab.io/llms-full.txt line 2071 gives Example 1 of Create a New Tab as 'A third-party delivery platform (e.g. DoorDash) submits a completed order... This is the most common pattern for delivery integrations', with a payload carrying both "phoneNumber" and "processor": "DOORDASH", and line 2015 'A customer must be associated with a tab'. Separately, guides/payment-flows carries a caution that reads almost like the claim: 'Creating a tab without an integration `customerId` does not make the tab anonymous. Supply another guest identifier accepted by the tab API, such as `phoneNumber`.' Note this quote is NOT in the cited spec -- it is in a different document -- and even taken at face value it is an INSTRUCTION TO THE INTEGRATOR about what to send, inside a page about choosing a checkout flow for customers the integrator itself signed in; it is not a report that marketplaces send it. The example number is the dummy '1233455678'. A worked example's placeholder cannot establish that DoorDash relays a reachable guest phone. POSITIVE COUNTER-EVIDENCE, which is why I did not merely trim this to a weaker partial. GoTab's documented route to a populated customer is an affirmative act the DSP path structurally cannot perform: /v2/verification/challenge 'Creates or resolves GoTab customer and triggers OTP challenge for verification', and rewards check-in by phone at the POS or CFD. GoTab's own KB states the default the other way -- a guest who does not check in 'should start a new tab and close it out anonymously, without checking in' (llms-full.txt line 7206). And the operator-side setup doc, which is the only document describing what a real DSP order looks like inside GoTab, tells the operator the connection needs 'a spare cell phone number. Google offers one free Google Voice number', records payments only as API_Integration, and never mentions a customer record. SCOPE: the public API is the integrator surface. The Stream leg is third-party and undocumented here, so the spec bounds what an integrator MAY send, not what a marketplace DOES send. A partial that shortfalls the conjunct carrying the claim's entire point, on a surface never bound to the subject, is self-refuting -- and the proposer says so themselves in their own caveat ('a verifier could reasonably hold the whole cell at unknown'). Not downgraded to no either: none of this enumerates and denies marketplace identity attach, so it is absence of evidence, which is unknown. GRADE: A would have been defensible for the spec's text but the proposition is not in the spec, so the grade question is moot. NO DOUBLE COUNT INTRODUCED, and one avoided: /integrations/customers (the Customers API) is already the evidence under guest-loyalty-unified-profile (partial/B), so accepting this cell on that same resource would have banked one API twice; delivery-3p-direct-integration (partial/B) and delivery-3p-reconciliation (no/B) already price in the Stream route. No gotab claim currently cites docs.gotab.io/specs or /graphql-schema.txt, so nothing is disturbed. SUBSTRING TRAPS CHECKED: 'handle' here is a contact handle (IntegrationCustomerInput.handle is documented 'The phone number for the customer'), not a social handle -- the proposer got that right; customerName/customerPhone in the spec are NOT tab or order fields and must not be read as marketplace identity -- all six occurrences belong to TabPass/tab-passes request bodies ('the name of the customer associated with the pass'), a spend-limit pass feature; and Tab.gotabOrder is a GoTab-ownership boolean, not a channel flag. The standing rationale's falsification test is unchanged and correct: an article or API note stating whether a Stream aggregator order creates or matches a GoTab customer. source
extensibility-middleware-compatibilityunknown, grade Fresolve-to-partialValue upheld, grade downgraded B -> C, and the citation re-homed. VALUE: partial is correct and peer-consistent. Peer convention is unambiguous — 1 of the 5 enumerated platforms scores partial (hungerrush partial/C 'Checkmate is the only aggregator ... correctly refuses the two-platform bar'; lightspeed partial/B; xenial partial/E), 2+ scores yes, and an off-list aggregator alone is still partial (odoo-pos partial/B on UrbanPiper). So Stream, not being among the five, cannot carry the cell to yes, and Otter alone cannot either. It is also not unknown: Otter's own index listing GoTab under 'Point of Sale' is positive evidence, which is exactly the surface the standing rationale named as dispositive ('resolved by a Deliverect or Chowly POS-support page naming GoTab'). DIRECTION: I checked Otter the same way the proposer checked Deliverect, and it passes both tests — the index category is 'Point of Sale' (the same index files Flipdish/Slice/ChowNow under 'Online ordering', so Otter does distinguish the directions), and the body says 'your GoTab POS system'. I confirmed all three Otter quotes verbatim from the 281 KB raw bytes, not from a summary: a fast-model read of the detail page asserted the REVERSE direction in its prose while its own quotes said the opposite, so the byte-level check was load-bearing. Deliverect's exclusion is independently confirmed from raw bytes ('Category Online ordering'). GRADE: B is too generous. The grade-B GoTab help doc evidences only Stream, which the claim does not enumerate; the sole enumerated platform is attested only by tryotter.com marketing copy, which is the integrator's own selling page. Peers grade this source class C or E (hungerrush C, olo C, xenial E, clover/revel/oracle E). C, with final_url moved to the page that actually carries the enumerated proposition. NO DOUBLE COUNT: delivery-3p-direct-integration (partial/B) already prices in the Stream route, but for the opposite proposition (absence of a first-party certified integration), and neither delivery-3p-injection nor extensibility-order-injection-api scores middleware endpoint status; the only new information in this cell is Otter. UPGRADE PATHS CLOSED: ItsaCheckmate names 28 POS and omits GoTab; Cuboh's GoTab URL is a 404 body (my initial census hit on 'cuboh' was the filename string, a false lead); Chowly's two cached files are byte-identical at 50,584 chars, i.e. a catch-all shell with 0 GoTab. PROCESS DEFECT FLAGGED: chowly.com is listed in data/robots-bars.json and was fetched during the proposal pass; it is not cited here and changes no outcome. Finally, my own zero-census was initially FALSE — the first tokeniser reported streamorders in 0 files while raw bytes showed it 3 times in an open page; the published zeros come only from the repaired run that passes four positive controls. source
inventory-shelf-life-expiryunknown, grade Fresolve-to-noRe-fetched both specs myself and re-parsed them structurally (the dump is not raw SDL - it is a '### TypeName' markdown census, and my first parser returned 0 blocks until corrected). The substance of the proposal is confirmed verbatim: 349 type blocks; Inventory 39 fields, VariantInventory 28, InventoryAdjustment 12, with created/archived the only date columns; and all three cited enums are exactly as quoted and closed. I upgraded the evidence rather than just accepting it - the proposer censused only the output types, while `InventoriesOrderBy`, `VariantInventoriesOrderBy`, `InventoryAdjustmentsOrderBy` and the three *Condition inputs are PostGraphile-generated one-entry-per-COLUMN enumerations, and they too carry no date beyond CREATED/ARCHIVED, which is what makes this a genuine census of the table rather than of the integrator surface. I also answered the bounding objection the proposal left open: the manager dashboard does not escape the model, because GoTab's own docs for those very screens (operator/menu-management/inventory-management/ and operator/manager-dashboard/reports-inventory-depletion/) enumerate Recount/Received/Loss/Theft, Par Level, Out-of-Stock Action, Untrack, and a report returning Sku/Product/Total Adjustments/Units - count-only, with no use-by field and no expiring-soon view. Decisively for this claim's wording, RECEIVED and PREP are themselves adjustment reasons, so the 'received and prepped items' path demonstrably has no date slot. Ran the traps tokenised myself and reproduced them: whole-token {expires:1} in the schema (the Product.manualDelayUntil prep-delay comment) and {expire:5, expires:6} in the REST spec (card expire_month/expire_year, OAuth expires/expiresIn); shelf/spoil/perishable/fifo/lot/batch/waste/freshness are zero in both, and the docs-corpus 'rotation' hit is automatic SCREEN rotation in a Guest Pay note. Two factual errors in the proposal are corrected in the amended note: the REST spec has 65 PATHS and 90 OPERATIONS, not '90 paths'; and 'no date property' is too strong - InventoryItem/InventoryProduct carry `updatedAt`, a row-modification stamp (the conclusion is unaffected, but the overstatement would not have survived a reader checking it). The note now also states the opsi bound explicitly, since this record scores recipe/costing capability as living in that separately priced module. Not a double count: the schema grounds a proposition distinct from inventory-native-not-partner and inventory-par-auto-suggest. Grade A stands - a generated schema and an OpenAPI spec are primary technical documentation. This overturns a standing verified unknown, and legitimately so: the two prior passes rejected `no` because the only candidate was a task-oriented KB walkthrough and an opsi marketing page, neither of which is a data-model census; docs.gotab.io now publishes one, which is exactly the class of source those verdicts said was missing. source

Sources

Every URL this record cites. 116 in total.