Inside DoorDash's Empire Week: What the SevenRooms and Deliveroo Deals Mean for Operators

It is 7:42 a.m. Pacific on Tuesday, May 6, and I am on the phone with an operator who runs eleven full-service restaurants across two coasts. He is calling me from the back office of the unit closest to his house, the one he still works Saturday doubles in when a GM goes on vacation. He has not had coffee. He has read two press releases. He is using a word I cannot print here to describe what just happened to his tech stack.

“My reservation vendor,” he says, “just got bought by my delivery vendor. And my delivery vendor just bought a European delivery company for almost three billion pounds. On the same morning. Before I finished my run.”

He is, as of this morning, both a SevenRooms customer and a DoorDash marketplace partner. He uses SevenRooms for guest profiles, reservations, two-way SMS, and the marketing automations that drive his weekday covers in a market where weekday covers do not happen by accident. He uses DoorDash for off-premise, for DashPass-driven incremental orders, and — in three of his eleven units — for in-house delivery dispatch through Drive. He has, in other words, been running the textbook hybrid playbook that every multi-unit operator has been running since 2022.

As of about 6:00 a.m. Eastern this morning, that playbook is a different playbook.

I want to be direct with you, because the takes are going to come fast all week and most of them are going to mush the two announcements together. They are two separate transactions and you need to think about them separately, then think about what they mean together. The $1.2 billion all-cash acquisition of SevenRooms is the one that changes your North American operations this quarter. The £2.9 billion acquisition of Deliveroo is the one that tells you what kind of company DoorDash intends to be for the next decade. Read together — and Tony Xu absolutely wants you to read them together — they are the day DoorDash stopped being a logistics company that sells you orders and started being an infrastructure company that sells you a stack.

So here is my contrarian thesis, the one I have been arguing with operators on the phone all morning, and the one I will defend through the rest of this piece: if you are a multi-unit operator, you have 60 days to rebalance your vendor relationships, not six months. Treating reservations, CRM, loyalty, and demand aggregation as separate procurement categories — the way every consultant deck still tells you to — is over. The vendors do not see them as separate anymore. Your competitors who move first will not see them as separate either. And the operator who waits until Q4 budget season to “evaluate the implications” is going to find that the implications evaluated him.

Let me show you the work.

What May 6 actually consolidated

Start with the facts, because the framing is going to drift fast.

DoorDash announced this morning a definitive agreement to acquire SevenRooms in an all-cash deal valued at approximately $1.2 billion. SevenRooms — for any operator who has somehow avoided it — is the New York–founded guest-experience and CRM platform that powers reservations, waitlists, marketing automation, and direct-channel orders for north of 13,000 venues globally. Its client list is not a list of independents. It is Marriott. It is MGM Resorts. It is Mandarin Oriental. It is Wolfgang Puck Fine Dining Group. These are the operators who write checks that move markets, and they write those checks specifically for the guest-data layer that SevenRooms owns.

In the same release window, DoorDash announced an agreed offer to acquire Deliveroo, the London-listed delivery marketplace, at 180 pence per share, an enterprise value of roughly £2.9 billion. Deliveroo gives DoorDash a real footprint in the UK, Ireland, France, Italy, Belgium, the UAE, Kuwait, Qatar, Singapore, and Hong Kong — markets where DoorDash had no meaningful surface area as of yesterday. The combined entity, post-close, will be a roughly 1+ billion order per year platform across more than forty countries. That is a Western-hemisphere delivery rival to Meituan, finally, after a decade of false starts.

And under both of those announcements, also today, DoorDash printed Q1. The print missed on revenue but did the thing the company has been telegraphing for two years on the cash line: adjusted EBITDA of $590 million, free cash flow of $494 million, on a marketplace gross order value approaching $23 billion for the quarter. That FCF number is the entire reason the company can write a $1.2 billion cash check for SevenRooms and a £2.9 billion offer for Deliveroo in the same news cycle without flinching. The funding mechanism for empire-building was built into the operating model two years before the empire-building started. We will come back to this.

Here is what these three data points consolidate, in plain English. DoorDash now operates, or will operate within twelve months pending regulatory clearance, the following four layers of your restaurant:

1. Off-premise demand and dispatch. The original DoorDash marketplace, plus Drive (the white-label logistics rails you already use for your own website orders), plus, soon, Deliveroo’s marketplace in Europe and the Middle East.
2. On-premise reservations and waitlist. Via SevenRooms.
3. Guest CRM, marketing automation, and the single guest record. Via SevenRooms.
4. Loyalty. DashPass, the SevenRooms loyalty product, and — read the Restaurant Dive coverage carefully — the explicit Tony Xu framing that the combined platform will tie DashPass benefits into in-store experiences.

That fourth one is the move. If you only read one line of the SevenRooms press release, read this paraphrase of Xu’s framing: DoorDash wants to turn your dine-in guest and your delivery guest into the same guest record, owned at the platform layer. The acquisition is not really about reservations. It is about the guest ID.

I had an operator on the phone at 9:15 a.m. Eastern who put it better than I have. “They didn’t buy a reservation system,” he said. “They bought my customer list. Or, more accurately, they bought the company that holds my customer list, and now I have to decide whether I still trust the lockbox.”

Why SevenRooms operators have weeks, not quarters, to decide

If you are running SevenRooms in your dining rooms today, you have three honest options. I want to walk through them as an operator would, not as an analyst would.

Option one: stay, and lean in. The bet here is that DoorDash makes the SevenRooms product better. More integrated. Cheaper, possibly, if the contract gets bundled with your marketplace fees. The dine-in guest gets tied to the delivery guest, your retention metrics improve, and the operator who stays first benefits most. There is a real version of this story. SevenRooms under DoorDash with $494 million in quarterly FCF behind it is going to ship product faster than SevenRooms as a standalone Series-D-stage company that had to pick its bets carefully. Anyone who has watched what Toast did to its product velocity post-IPO knows what capital does to a roadmap.

Option two: stay, but firewall. Keep SevenRooms for the front-of-house workflow your team already knows. Renegotiate your contract — and we will get to the contract language in a minute — to put hard limits on what guest data flows back to the parent company’s marketplace side. This is the option I expect most of the Marriott-tier customers to take, because their legal departments will require it. It is also the option most multi-unit independents will land on, because retraining a host stand on a new platform during peak summer is not a thing operators do casually. The lift is real; switching reservation systems is an 8 to 12 week project per concept if you do it right.

Option three: leave. Migrate to OpenTable, Tock, or Resy in the next two to three quarters. This is the option being whispered loudest by the operators I spoke with this morning who do the largest delivery volume. Their reasoning is straightforward and worth hearing: “If I am going to have a fight with DoorDash about take rates in 2026, I do not want them to also own my CRM during the fight.” That is a procurement-leverage argument, not a product argument, and it is the correct argument.

The reason this is a 60-day decision and not a six-month decision is contract timing. Most enterprise SevenRooms contracts I have seen run on annual or biannual cycles, and the renewal windows that hit in Q3 are going to be the first ones that test the new ownership reality. If you wait until your renewal lands on your desk to start the conversation, you have already lost the leverage window. The leverage window is now, while the deal is freshly announced, while the integration is theoretical, and while SevenRooms’s account managers still have authority to make concessions without checking with Tony Xu’s team.

There is also a personnel dimension nobody is talking about today and everyone will be talking about by August. SevenRooms’s senior product leadership and its NYC engineering team are going to get reorged. Some will stay, some will leave. The account manager who has been a real partner to your operation for three years has a non-trivial probability of being on the other side of a LinkedIn announcement by Labor Day. If your retention strategy with SevenRooms has been built on individual relationships, those relationships have a half-life now.

This is the part where I want to be honest about my own bias. I am not a “rip-and-replace” person. I have written, repeatedly in this column, that switching costs in operator tech are systematically underestimated by people who do not have to retrain a 4 a.m. opener. But I have also watched what happens when a platform vendor’s incentives quietly shift away from the operator and toward the platform’s own marketplace side. The first six months feel fine. The second six months, the API access starts narrowing. The third six months, the export tools “get refactored.” This is not a hypothetical pattern. It is what every operator who lived through the Grubhub-and-Seamless era can tell you happened to their direct-channel order data between 2017 and 2020.

So: leverage now. Decisions by end of Q3. Migration plans, if you choose option three, scoped before Q4 budgeting starts.

The Deliveroo half of the story, decoded

I want to spend less time on Deliveroo than on SevenRooms, because for the typical North American operator reading this, Deliveroo is the geopolitical context, not the operational one. But the geopolitical context matters more than people are admitting today.

What DoorDash bought, for £2.9 billion, is not really Deliveroo’s order volume. Deliveroo’s order volume is fine but unspectacular, and the company has been in a years-long fight to get its UK take rates and rider economics to a place that satisfies public-market investors. What DoorDash bought is European optionality. It bought the right to put DashPass — or whatever the unified post-merger loyalty program ends up being called — in front of European consumers who have, until today, never heard of DoorDash. It bought the right to compete with Just Eat Takeaway, Uber Eats Europe, and Glovo on the same continent at the same scale, instead of the Wolt-shaped corner of Europe that DoorDash has been working in since the 2021 acquisition.

For the multi-unit operator with European units — and I know more of you read this column than I would have guessed five years ago — the immediate question is whether your Deliveroo marketplace contract still means what you think it means. The answer, today, is yes. Deals of this size close on a six-to-twelve-month timeline at minimum, and that timeline is materially extended by UK and European regulatory review. Nothing changes in your day-to-day until at least Q4 2025, and probably not until 2026.

For the multi-unit operator with only North American units, the Deliveroo deal matters for one reason: it tells you that DoorDash’s capital allocation strategy is now empire-building, and empire-builders need to monetize the empire. The Deliveroo acquisition is not free. The £2.9 billion does not come out of thin air. It comes out of free cash flow that, in the alternate universe where DoorDash did not buy Deliveroo, could have gone to lower take rates, better restaurant economics, or stockholder returns. You should not expect take rates to go down in 2026. You should plan, in your 2026 P&L modeling, for them to stay flat in the best case and tick up in some markets in the realistic case.

This is what the Business Wire announcement does not say but every operator should read between the lines of. The pitch to public markets today is global scale and consumer ubiquity. The funding for that pitch, in the medium term, comes from the same place it has always come from: the margin pool between what consumers pay and what restaurants get.

How OpenTable, Tock, and Resy just got a competitive lever they did not have yesterday

Here is the part of the story I think the day-one coverage is going to underweight, and that operators should pay attention to most.

OpenTable, Tock (owned by American Express), and Resy (also owned by American Express) — the three competing reservation platforms — woke up this morning with a sales pitch they did not have yesterday. The pitch writes itself: “Your reservation system should not be owned by the company that runs your largest off-premise demand channel. Our reservation system is not owned by them. Switch.”

That pitch is going to land hardest with three operator segments: the fine-dining segment that has historically been over-indexed to Tock and Resy; the hotel-restaurant segment that is over-indexed to SevenRooms and that has corporate procurement departments allergic to vendor concentration; and the high-end independent multi-unit operator who has built a direct-channel business and views their guest data as a strategic asset rather than a cost center.

I have been told — and I cannot get on-the-record confirmation for this until later in the week — that OpenTable’s enterprise sales team had a Tuesday morning all-hands at 8:00 a.m. Eastern to reposition their pitch decks. Tock’s parent at American Express is going to read this deal as a green light to invest more aggressively in Tock’s product roadmap, because the competitive moat just got more interesting. Resy is sitting on the same opportunity.

For the operator, this is unambiguously good news in the short term. Three vendors are now incentivized to compete harder for your business than they were yesterday. If you are inside a SevenRooms renewal cycle this year, you should be taking OpenTable, Tock, and Resy meetings even if you have no intention of switching, because the existence of those meetings is itself a leverage event with your incumbent. In a piece we publish later on the loyalty layer, I will go deeper on how loyalty program portability becomes the next battleground here — the operators who own their loyalty IDs at the POS layer rather than the reservation layer come out ahead — but for today, the takeaway is simpler: the reservation market just got more competitive. Use it.

The capital story behind the empire

I promised earlier I would come back to the Q1 cash numbers, because they are doing more analytical work than the headlines about the deals themselves.

DoorDash put up $590 million in adjusted EBITDA and $494 million in free cash flow in a single quarter. Annualize that and you have a roughly $2 billion FCF business. That is the number that lets a CFO sign off on a $1.2 billion all-cash acquisition and a £2.9 billion offer in the same week without diluting shareholders and without taking on punitive debt. The company is funding empire from operations.

Compare this to where DoorDash was three years ago, when the post-IPO honeymoon had ended and the company was getting hammered for unit economics that did not pencil out at scale. The cash conversion machine that the company built between 2022 and today is the entire reason this morning’s announcements exist. If Q1 EBITDA had been $290 million instead of $590 million, this is a very different press cycle.

The strategic implication for operators is the one I led with: the funding mechanism is now empire-building. The capital is going to be deployed into more acquisitions — I would not be surprised to see a POS adjacency move in the next eighteen months, and I would not be shocked by a payments or BNPL move on a longer horizon — and into product integration across the now-larger surface area. The capital is not going to be deployed, on any meaningful scale, into making the unit economics of being a restaurant on DoorDash’s marketplace fundamentally better for the restaurant.

That is not a moral judgment about DoorDash. It is a structural observation about what publicly-traded growth-stage platform companies do with cash. Toast did it. Olo did it on a smaller scale. Square did it for a decade. The pattern is consistent. The operator’s job is not to be angry about the pattern; it is to plan around it.

What an operator should write into their Q3 contracts

This is the practical section. I want to give you specific contract language to insist on between now and the end of Q3, whether you are renewing with SevenRooms, renewing your DoorDash marketplace agreement, signing with a SevenRooms competitor, or negotiating with anyone in the broader CRM and loyalty stack.

Data portability clauses with teeth. Your guest data — every email, every reservation history, every dietary preference, every visit-frequency score — must be exportable in a machine-readable format within 30 days of a written request, at no additional fee, throughout the term of the contract and for at least 12 months after termination. If your current contract does not have this, you do not have a CRM; you have a hostage situation. This was always true. It is more true today than it was yesterday.

Cross-platform data-sharing restrictions. If your reservation vendor is owned by, or becomes owned by, a marketplace company that you also use, you need an explicit, opt-in-only clause governing what guest data flows between the two sides. Default-off, with audit rights. The Marriott legal team is going to insist on this for every venue under the Marriott umbrella. You should too.

Take-rate protections, even where you cannot get rate locks. I am realistic; you are not going to get a multi-year take-rate lock from DoorDash, and you should not pretend you can. But you can negotiate notice periods — 90 days minimum before any take-rate change — and you can negotiate audit rights on the calculation of effective rates after promotions, commissions, and bundle pricing are applied. The effective rate is the only rate that matters, and it is the one that drifts.

Off-ramp language. Every contract you sign in 2025 should have explicit, fee-free off-ramp language for the case where the counterparty is acquired, materially changes its product, or materially changes its pricing. This is standard in enterprise SaaS and historically less standard in restaurant tech. Push for it. The legal cost of getting it into your contract is trivial compared to the operational cost of being trapped.

Loyalty-data ownership. If your loyalty program is operated by a third party — and increasingly, if your reservation platform is operating loyalty for you, it is — you need explicit ownership of the loyalty IDs, the points balances, and the underlying transaction data. The framework piece we publish later in this column on operator-side data infrastructure goes deeper on the technical architecture of owning your own IDs, but the contract-language version is simple: insist that the data lives on systems you control or can extract on demand, not on the vendor’s systems with promises of future export.

I want to flag that none of these clauses are exotic. They are standard in enterprise SaaS for industries with longer maturity curves than restaurant tech. The reason they are not yet standard in our world is that the procurement function in most multi-unit operators is built around food and beverage cost, labor, and real estate, and the tech contracts get signed by an operations lead who is not a procurement specialist. That has to change, and the operators who change it first will, again, come out ahead.

The kitchen-side question nobody is asking yet

There is a thread I want to pull on that I have not seen anyone else pull on today, because it is one quarter out from being obvious and two quarters out from being urgent.

If DoorDash now operates the demand layer, the reservation layer, the CRM, and the loyalty program, the natural next platform integration is the kitchen. Not the POS — DoorDash already integrates with every major POS through the order injection layer — but the actual kitchen production system. The kitchen display, the prep forecasting, the labor scheduling against forecasted demand. The dot that connects “this guest just walked in” to “this guest ordered delivery last Tuesday and is likely to order the chef’s special tonight” to “the line needs to be staffed for a 7 p.m. push.”

Operators who have invested in automated kitchen infrastructure already have a head start on the data architecture that makes this kind of cross-channel forecasting tractable, as our later operator case study on automated kitchens argues. For everyone else, the lead time on getting your kitchen production data into a state where it can be merged with guest-side data is twelve to eighteen months of project work, and the project does not start itself.

I am not saying DoorDash is going to acquire a kitchen-tech company in 2025. I am saying that the strategic shape of the company they are building points there eventually, and the operator who treats kitchen-side data as a separate domain from guest-side data is going to be on the wrong side of the integration when it happens.

Operator takeaways

If you read nothing else this week, read this:

• Treat May 6 as the day vendor categories collapsed. Reservations, CRM, loyalty, and demand aggregation are now one procurement category, not four. Procure accordingly.
• If you are on SevenRooms, decide your posture by end of Q3 — stay, firewall, or leave. Each option has a defensible case; defaulting to “wait and see” is the only indefensible one. Start the OpenTable, Tock, and Resy meetings this week even if you intend to stay.
• Rewrite your data clauses now, not at renewal. Data portability, cross-platform sharing restrictions, audit rights, and off-ramp language are non-negotiable in any contract you sign between today and December.
• Model 2026 with flat-to-rising take rates. The Deliveroo deal is funded out of the same cash pool that, in an alternate universe, could have lowered your marketplace economics. Plan as though it will not.
• Own your guest IDs at a layer you control. Whether that is a POS-native customer record, a dedicated CDP, or a homegrown identity layer, the operators who emerge from the next 24 months with leverage are the ones whose guest data does not live on a vendor’s servers behind a vendor’s API.

The eleven-unit operator I started this piece with hung up the phone at 8:11 a.m. Pacific. His last sentence to me was the one I have been turning over all morning: “I think we just found out who our landlord is.”

He was being a little dramatic. He is also not wrong.

— Priya files The Operator. Tips: tips@tabletransfers.com.