Desk Review: Toast IQ at One Hundred Days — What the Four-Month Mark Actually Looks Like From the Outside

The coffee was on the second pour when I did the arithmetic. Toast IQ went generally available on October 29, 2025 — the Sous Chef rebrand Hana wrote up the same week. From October 29 to today, Wednesday, February 4, 2026, is ninety-eight days. Round up: a hundred. The four-month anniversary lands in the last week of February, which puts Toast IQ’s first proper birthday — the one Wall Street will write about — exactly eight days after Toast’s Q4 print on February 12.

That timing is not an accident. The Q4 call is going to be the moment when Aman Narang or Elena Gomez stands up and tells everyone how many locations have used the assistant, how many queries have run through it, and what the next eighteen months look like. We will write that piece next Wednesday with the transcript in hand. This piece is the before piece — the one that says what we already know on February 4, with the assistant a hundred days into GA and Toast IQ multi-unit twelve days into beta, and asks what the public surface actually tells us about whether the product is working.

The short version. Toast IQ has shipped more product in the last hundred days than most vendors ship in twelve months: five named modules through January, the multi-unit beta live, a paid SKU for the first time, the handheld scanner updated, a Google-and-Meta marketing integration. What we don’t have on February 4 is the operator outcome. The two answers Toast is most likely to volunteer on February 12 — percentage of locations that have used the assistant and cumulative query count — are vanity baselines. Neither is the question we’d ask if we’d been in the audience.

Methodology disclosure, module-by-module, questions for the call.

Methodology disclosure (the part Vibe Check requires up front)

A proper Vibe Check is a six-operator panel running the named product against a structured weekly log for ninety days, with three interview checkpoints. We have not cleared the bar because the multi-unit version — the part that matters for any operator above one location — has been in beta for twelve days, since January 23. There is no operator on this who has thirty days. The handful who do have a hundred days are running the per-location version that landed on October 29, which is a different product than the one Toast is selling on February 12.

This is therefore a Desk Review, the format I lean on when a release is too fresh to panel — the same shape I used for the broader Toast suite read in May and the January retailer release. I work from public materials and say what I think module by module, with confidence levels and the interpretive moves flagged.

The sourcing on February 4 is narrower than I would like. The community board has January 2026 product update notes; the February 2026 update thread is up but the public render is blocking external fetchers, so the canonical text we have is the one operators emailed us through the weekend. The trade press has the Restaurant Business drive-thru piece and the Digital Transactions assistant write-up. Toast’s January 23 multi-unit press materials and the Toast IQ Grow announcement round out the corpus. The Q4 transcript does not exist yet — the call is February 12 — and any number you read attributed to that transcript today is either fabricated or sourced from a different document.

Where I quote a stat I can only attribute to Toast, I mark it as a vendor claim. Confidence on each verdict is HIGH / MEDIUM / LOW with the same scale used since the May suite review: HIGH means a panel would likely confirm it, MEDIUM means a panel could easily flip it, LOW means treat the verdict as a hypothesis.

What has shipped through January 31, 2026

Five named modules and a platform move underneath them. The numbering tracks the order they hit operators.

1. Toast IQ multi-unit lookup (beta) — the assistant now answers questions across an operator’s locations. Beta opened January 23.
2. AI Invoice Scanning — inside Toast IQ; vendor invoices photographed and parsed without manual entry.
3. Toast IQ Grow — a $499/month paid SKU pairing the assistant with the marketing and advertising stack. First Toast IQ product priced separately from the platform fee.
4. Toast Advertising on Google and Meta — campaign creation and optimisation surfaced through Toast IQ for paid acquisition.
5. Toast Go 3 — barcode scanning on the handheld terminal. Hardware story, workflow consequences.

Underneath the modules: Toast IQ is no longer a per-location product. It is a multi-location product with single-location accounts as a special case. That structural move — which Maya read on January 23 as the real Toast IQ story — is the thing the February 12 call is going to organise around, and it is the thing every module from here is going to either lean on or sit beside. Now the verdicts.

Module-by-module verdict

Toast IQ multi-unit lookup — verdict: keep the beta seat. The hundred-day version of this product is not the product Toast is going to be selling at the end of Q2. (Confidence: MEDIUM)

The headline. Through January 22, Toast IQ was a per-location product. From January 23, in beta, the assistant looks across multiple locations on the same operator account. Which of my five stores is overstocked on the seasonal SKU? Where is the price mismatch between Store 3 and Store 7? That is what the lookup does.

The public reading on adoption is the number the trade press has been quoting since January: more than half of Toast’s 164,000 locations have used Toast IQ at least once since GA, with cumulative query volume in the millions. We cited it in the rebrand piece and again in the January desk review; both times I flagged that “used” is a low bar and cumulative queries is a vanity number. The directional read — operators do open the chat — is the prerequisite for the multi-unit version landing rather than getting ignored. That prerequisite is met.

Freshness problem worth naming. By February 12, the multi-unit beta will have eighteen days of usage. By April 1 — the date I would put on a panel — ten weeks. Whatever Toast announces on February 12 about Toast IQ adoption will be true mostly of the per-location product. Read every figure with that frame.

Omri Traub’s “right hand” line — “operators have an AI-powered ‘right hand’ partner that understands the flow of inventory and pricing as naturally as a store manager does” — is still the most useful sentence Toast has put on the record about the multi-unit pitch (Source: Digital Transactions). The structural read I named in the January desk review holds: multi-unit lookup is, in shape, a chain-level operator console. Toast didn’t pitch it that way. I am, because the lookup moves the unit of analysis up a level.

The risk — and why I’m at MEDIUM — is the one we flagged in the suite review: Toast IQ sees what Toast sees. A multi-unit lookup answering cross-store inventory questions confidently while missing third-party distributor pricing or loyalty data in another system is more dangerous than the single-store version, not less. Confidently-wrong cross-store recommendations scale. The most important number Toast could publish at four months is not the percentage of locations that used the assistant. It is the percentage of recommendations operators acted on, broken down by integration depth. I will bet against that number making the slide.

Reach for it if: you run three-plus locations on Toast, your data layer is mostly on Toast, and your cross-store reporting is a Tuesday spreadsheet ritual. Even in beta the lookup is a step-change. Hold off if: you’re single-location, or you run a multi-unit footprint where two-plus systems carry materially different pieces of the operation. Wait until the assistant integrates them.

AI Invoice Scanning — verdict: keep, and watch the error rate. (Confidence: MEDIUM)

The module operators will feel the most through Q1. AI Invoice Scanning lets a manager photograph or upload a vendor invoice and have Toast IQ extract the line items, vendor, totals, and dates without manual keying. The hand-key step is one of the worst recurring evenings the modern POS has failed to fix, and the demo is genuinely strong.

A hundred days in, the operator commentary in the trade press and on the community board is consistent on two things. The floor cases work — printed invoices from large distributors with clean OCR characters parse cleanly. The ceiling cases miss — handwritten deductions, pen marks, distributor-specific abbreviations, and the occasional dot-matrix print still require manual reconciliation. None of that is surprising. All of it is what a panel would catch in week two.

What I have not seen Toast publish is a parse-accuracy number. The press release frames invoice scanning as eliminating manual data entry; the community notes describe the workflow; the trade press repeats the framing. None of them tell an operator what percentage of invoices land cleanly without a manager touching them. If the February 12 call mentions invoice-scanning accuracy, the bar to clear is north of ninety percent on clean prints. If the call mentions “hundreds of thousands of invoices processed” without an accuracy figure, treat the metric as marketing.

Reach for it if: you take fifteen-plus vendor invoices a week and your current process is photograph-then-rekey. The floor savings are real. Hold off if: your invoice data must be exactly right for tax or audit reasons and you don’t have a reconciliation step. Add the step or wait for a vendor-published accuracy number.

Toast IQ Grow ($499/month bundle) — verdict: still too new to keep, still too revealing to ignore. (Confidence: LOW)

The module the financial-strategy crowd should read most closely. Toast IQ Grow is the first Toast IQ product packaged as a paid SKU separate from the platform fee — $499/month, bundling the assistant with the marketing and advertising stack. Up to now, Toast IQ has lived inside the core platform price. This is the first piece of that closing, as I flagged in the January desk review.

The financially-interesting question is the attach rate. Toast has not published one. The February 12 call may give us a number — Toast loves an attach-rate slide, and Grow justifies the disclosure if the number is good. Expect language like “thousands of operators have already adopted Toast IQ Grow” without a denominator. The denominator that matters is the percentage of Toast’s locations that also run an active marketing programme — the addressable base for Grow, not the full 164,000.

Pricing read — interpretation — unchanged from January: $499/month is the price you set when you expect operators to renew within twelve months. High enough to fund product investment, low enough that loss aversion on cancelling is real. Toast is pricing Grow as a commit, not a trial.

The vendor lift numbers in early commentary — including a return-on-ad-spend figure cited as “up to 8x” in pilot deployments — are vendor pilot numbers, with the caveats vendor pilots always require: no published control, no operator-side measurement methodology, no panel against a baseline. Treat the 8x figure as a vendor claim, not an outcome.

Reach for it if: you have no marketing programme today and $499/month is less than your fully-loaded cost of a five-hour-a-month freelance marketer. The agent will give you something. Hold off if: you already run a marketing stack you trust. Wait for an against-baseline number.

Toast Advertising on Google and Meta — verdict: it is real and it is integrated. The independent measurement is still missing. (Confidence: LOW)

Related to Grow but worth pulling apart. Toast Advertising on Google and Meta is the campaign-creation-and-optimisation layer that sits inside Toast IQ. An operator briefs a campaign in plain language — “Wednesday wine night, promote to lapsed guests within five miles, $200 budget” — and the assistant builds the audience, drafts the creative, and pushes the campaign with optimisation toggles.

A hundred days into Toast IQ’s life, the operator commentary on the marketing side is sparse. The community board has Toast’s walkthrough copy, the trade press has Toast’s framing, and the operator panels that exist — including the May suite review — are still talking about the pre-Grow marketing stack. What we don’t have on February 4 is operator-side reporting on whether agent-built campaigns converted better than operator-built ones at the same budget. That is the panel question.

What I would watch for on February 12 is whether Toast volunteers a Google-and-Meta-specific campaign volume number. If the call gives us “X thousand campaigns launched through Toast IQ in Q4,” it is a usage number, not an outcome. The outcome number is the lift over an operator’s prior baseline, and Toast is unlikely to put a lift number on a slide deck unless the lift is comfortably positive.

Reach for it if: you would otherwise have no campaigns running through Google or Meta in Q1 — anything the agent builds is a floor. Hold off if: you have a marketing partner running campaigns against a measured baseline. Don’t swap a measured programme for an unmeasured one without a published comparison.

Toast Go 3 — verdict: keep, and watch how barcode lands on the floor. (Confidence: MEDIUM)

The hardware module that sounds like a hardware story and is actually a workflow story. Toast Go 3 — the handheld terminal — picked up barcode scanning as a standard capability in January, which means the device now does receiving, count, and sale-floor lookup in one workflow rather than three.

Toast Go 3 is the module operators on the community board have been most consistently positive about — the hardware does an obvious job, the workflow change is concrete, and the AI surface is in the background, not the foreground. The thing to watch is the integration story: a Toast Go 3 in the hand of a receiving manager is more valuable when multi-unit lookup can tell her which other store needs the case she’s about to mis-flag as overstock. Cross-store routing is the future of the device. The present is receiving in one workflow, and that is enough.

Reach for it if: you have receiving, count, and sales-floor lookup as separate workflows. The consolidation alone is worth the upgrade. Hold off if: your existing handheld stack works and your team doesn’t do receiving on it.

The platform move underneath the modules

The structural move I keep coming back to: Toast IQ has crossed from a per-location product into a multi-location product, and the question for Q2 is whether the per-location adoption transfers.

There is a real argument that it doesn’t, automatically. The per-location use case is a GM asking “why did my Friday close run light?” on the terminal at the end of a shift. The multi-unit use case is a regional manager asking “which of my eleven stores is the outlier on Friday close?” from a back-office screen Monday morning. Different products, different operators, different physical rooms. Half of locations opening the chat doesn’t mean half of regional managers will.

Interpretive read — flagged — Toast is going to count multi-unit adoption against the same 164,000-location denominator as per-location adoption, and the headline number on February 12 will therefore be a per-location number dressed in a multi-unit announcement. Watch for whether Toast publishes a separate multi-unit number against a denominator of operators — three-or-more-store operators in particular. That is the number that tells us whether the chain CIO console pitch is landing. I will bet against it being on the slide.

What’s coming that we already know about

Two adjacencies are worth flagging before the forward-look.

Drive-thru. The Restaurant Business piece on Toast’s drive-thru plans lays out the strategic move: Toast intends to launch a dedicated drive-thru product, with multi-lane support, in 2026. Aman Narang is on the record — “Launching a drive-thru product is going to open up that market in a much bigger way” — and Toast’s 2023 acquisition of Delphi Display Systems gave it the digital-signage and software foundation. The drive-thru market is structurally different from the per-location restaurant POS Toast has been winning: multi-lane configuration, KDS integration, voice-AI for order capture, enterprise sales cycle. Toast IQ will sit somewhere in that stack — the natural fit is voice — but the drive-thru story is its own roadmap, and February 12 is going to talk about it as a separate workstream.

Enterprise wins. The same piece names Applebee’s, Firehouse Subs, Papa Murphy’s, and pegs Toast’s 2025 net-new locations at thirty thousand against a base of 164,000. Whether those enterprise accounts use Toast IQ in 2026 is the leading indicator on chain adoption. None of the named brands has been disclosed as a multi-unit beta participant in public materials. Add it to the call questions.

What to watch on the February 12 call

The piece I will write next Wednesday is the transcript piece — the one that reads what Toast actually said against the public surface this Desk Review has mapped. Before then, the questions a careful operator should hold in mind:

1. Multi-unit adoption against what denominator? Adoption against 164,000 locations is a per-location number. Against three-or-more-store operators, it is a chain number. Different questions.
2. Recommendation-acted-on rate, broken down by integration depth. Toast-only data versus non-Toast data should look different. One number for both is hiding the distinction.
3. Parse-accuracy for AI Invoice Scanning. Above ninety percent on clean prints is a real claim. “Hundreds of thousands of invoices processed” without accuracy is marketing.
4. Toast IQ Grow attach rate, against what addressable base. “Thousands of operators” against 164,000 is one story. Against the smaller base of locations running an active marketing programme is the more interesting one.
5. Multi-lane drive-thru launch window, and what does Toast IQ do inside that stack? If the answer is “Toast IQ is the foundation of everything we ship from here,” it is positioning. If it is “Toast IQ powers voice ordering and drive-thru menu intelligence, the rest is conventional,” it is a roadmap.
6. A named enterprise Toast IQ multi-unit reference. Named chain on the slide is the leading indicator. “A top-25 QSR brand” is deferred.
7. Buyback expansion, and size. A meaningful buyback against the AI investment story is the “we can do both” signal. Size, if it comes, is information.
8. A Toast IQ revenue line broken out from platform. First time Toast breaks it out is the first time the AI surface has its own P&L disclosure.

Some questions will get answered. Most won’t. The ones that don’t are the ones a panel — ours, in Q2, against a cohort with sixty days on multi-unit — will answer instead.

The one paragraph I want operators to take into the call

Toast IQ is a hundred days into general availability, twelve days into the multi-unit beta, ten weeks into having a paid SKU, and shipping faster than most vendors in the category. That is a real product moment. The adoption numbers Toast will name on February 12 are true — but they are the wrong numbers to grade the product on, because they are usage, not outcome, and they are mostly per-location numbers in a moment that has moved to multi-unit. Listen for the outcome disclosures — recommendation-acted-on rates, parse accuracy, attach rate against the right denominator, named enterprise references — and discount the usage disclosures by a healthy margin.

A footnote on what we don’t know. The February product update thread is public but not fully retrievable through our normal channels this week; we worked off operator screenshots. The full text will land through the month, and any module under-read here will get a follow-up. Same goes for the SEC Q4 exhibit — published with the call on February 12, and we will read it line by line.

The Q4 call is in a week. The four-month anniversary lands the week after. The piece that grades Toast IQ at four months is the one I will write on February 13 with the transcript open. This is the piece that says what is already on the floor on February 4, and what to listen for when Aman Narang stands up to tell the market how the first hundred days went.

See you next Wednesday.

— Sofia leads vendor reviews for TableTransfers. Tips: [email protected].