The OpenTable Squeeze, Heading Into Q4

Monday morning, 8:14am at a counter in the Mission, and the operator I’m having coffee with is asking the same question three different operators asked me last week: should we be looking at Resy again? It’s an annual-renewal question. Their OpenTable contract is up in March. Booking Holdings reports Q4 on February 19. And the operator Slacks I lurk in have shifted from “OpenTable is fine” to “OpenTable is doing the thing again” — tightening the rules around what restaurants can do with their own reservation inventory.

Let me say what I think the shape of this story is, going into earnings, and then walk through the part of the math that doesn’t move.

The Soo-era ledger, five and a half years in

Debby Soo has been CEO of OpenTable since August 2020. That’s a long enough tenure to grade. The bull case her team would write: OpenTable came out of the pandemic with a network of 60,000+ restaurants globally, shipped Concierge in July, was the first restaurant tech partner for OpenAI’s Operator agent, and is currently the booking layer under most of the AI-discovery surfaces being built. The brand survived. The seat count grew. The strategic position inside Booking Holdings looks more central than it has in a decade — Marco walked through that math in The Pass three weeks ago.

The bear case the operator Slacks would write: the same five and a half years gave us a deeper data moat that runs in only one direction. Restaurants put data in. Getting it back out — into a CRM, into SevenRooms, into a marketing tool the restaurant actually owns — keeps getting harder. The 2019 episode is the canonical example. OpenTable updated its client agreement to bar restaurants from sharing diner data with competing platforms without paying fees, specifically targeting SevenRooms’ integration via restaurant login credentials. CEO Steve Hafner framed it as privacy. SevenRooms CEO Joel Montaniel framed it as restaurants being prevented from “operating freely.” Both can be true. The structural fact is the one that matters: when the booking platform owns the data wire, the platform decides who else gets to touch it.

That episode is six years old. The pattern hasn’t changed. The branding around it has gotten more polished.

What the chatter is actually about

I’m not going to retail-flog rumours. But the same phrase shows up in three separate operator conversations this month: primary system of record. Multi-location groups are reading new contract language — or being told they will be reading new contract language at renewal — that nudges OpenTable from “one of our reservation channels” to “the one our website points at and the one our inventory ultimately reconciles to.”

I don’t have a contract redline in hand as of this writing. (Interpretation flag: the rest of this paragraph is my read, not a sourced claim.) What I think is happening is a slow tightening of the same posture OpenTable has held since 2019 — the platform wants to be the system of truth for inventory, and “system of truth” lands very differently if you’re a single independent than if you’re a 14-unit group with a real CRM strategy. Later reporting will either confirm a concrete contract change or show this was rumor cycle ahead of earnings. Either way, the question on the table — should we switch? — is real.

The switch-cost math, which has not changed

Here is the part of the conversation that has gotten worse, not better, over five and a half years. (Interpretation flag: this is my framework, not industry-standard.)

The naive switch-cost math is: new platform’s monthly fee minus old platform’s monthly fee, times 12, plus an implementation week. If that were the real number, half the OpenTable network would have left for Resy or SevenRooms by now. They haven’t. The real switch cost is in four buckets the spreadsheet doesn’t see.

One: the inbound demand graph. OpenTable’s site, app and partner surfaces — ChatGPT, Operator, hotel concierge desks, Amex card portals — produce a meaningful share of cover volume for most restaurants on the platform. Switching means a transition period where that inbound goes dark. Operators who tell me they “tested” Resy usually mean they ran it in parallel for a quarter and renewed OpenTable because the cover delta scared them.

Two: the diner-side address book. Returning OpenTable diners have a profile and a one-tap flow. The new platform’s user has to find you, install a thing or sign up, and rebook. The drop-off is invisible on the operator’s dashboard. It shows up as soft Tuesdays six months later.

Three: the integration tax. POS, CRM, gift cards, loyalty — every middleware wired to OpenTable webhooks has to be re-wired. Some integrations don’t exist on the alternative platform yet. The 90-day migration is rarely 90 days.

Four: the staff retraining. Operators under-budget this one. The host stand has muscle memory. Two weeks of slower seatings and missed VIP flags is a real revenue line.

The switch only pencils when the contract delta is large or the strategy depends on owning the guest data outright. For most multi-units, it’s the second. They aren’t switching because Resy is cheaper. They’re switching because they’ve decided the data is the asset and they can’t run the marketing program they want while OpenTable owns the wire.

What to watch on February 19

Three things I’ll be listening for on the Booking Holdings Q4 call.

One: forward language on OpenTable contract terms. Booking Holdings does not usually narrate OpenTable contract structure on earnings calls. If they do this quarter — even at the level of “evolving our partner terms to support our AI distribution roadmap” — that’s the tell. (Interpretation flag: my read.)

Two: restaurant count. OpenTable has been quoting a global network number near 60,000 for the last several quarters. Net adds — not gross — is the number that tells you whether the Slack chatter is a churn signal or just chatter.

Three: Concierge and Operator metrics. Marco’s parent-company-strategy frame is the right one. The AI features are not a side project — they’re the reason the data moat matters more in 2026 than it did in 2019. A concrete demand-intercept number tied to OpenTable AI surfaces would be the real moat update.

If you’re the operator I had coffee with this morning: don’t sign the renewal yet. Read your March renewal letter line by line when it lands, get any new “primary system” language to your lawyer, and run the four-bucket math against your own group honestly. Resy is a real option. SevenRooms is a real option. The platform you’re currently paying is still, on most days, the rational answer.

The squeeze is real. The exit is harder than the chatter makes it sound. Both have been true since 2019, and both will still be true after February 19. The interesting question is which way the gap is widening.

— Maya covers restaurant tech for The Pass. Tips: maya@tabletransfers.com.