Uber agreed to pay $2.3 billion in cash for ezCater on October 6, 2026. DoorDash closed SevenRooms at $1,152 million and Deliveroo at $3,724 million in 2025. Thoma Bravo took Olo private for $2 billion. Catering, reservations, delivery, digital ordering. Four channels restaurants thought they controlled, all bought inside eighteen months.

Which Channels Changed Hands, and for How Much?

Five acquisitions, four buyers, close to $9.6 billion. Each one moved a restaurant channel from an independent vendor to an owner with its own commercial interest in your orders.

DoorDash's numbers sit in its own 10-Q. SevenRooms closed June 13, 2025, at $1,152 million in consideration, $902 million of it cash. Deliveroo closed October 2, 2025, at $3,724 million. Goodwill booked on the pair: $890 million and $1,985 million.

Uber announced ezCater on October 6, 2026, at $2.3 billion cash. The deal is expected to close in the coming months. Nihad Rahman stays on as chief executive.

Thoma Bravo signed for Olo on July 3, 2025, at $10.25 a share, roughly $2 billion all cash. Olo runs digital ordering for more than 750 restaurant brands. A private equity firm now owns that plumbing.

American Express took the reservation route earlier. Resy in 2019, Tock for $400 million in 2024, the two merged into one network in August 2026.

ezCater spent eighteen years as the alternative. Stefania Mallett and Briscoe Rodgers founded it in Boston in 2007, two years before Uber Eats existed. The pitch was simple. Sell catering without a delivery marketplace taking a cut. HNGRY called it the anti-DoorDash.

It just sold to the other one.

What Does Uber Buying ezCater Mean for Restaurant Economics?

Bigger orders, cheaper commission, and a new owner deciding how long both of those last. ezCater's average order runs about $430. A typical delivery order sits near $30.

The commission gap explains why restaurants tolerated the platform. ezCater's take rate runs roughly half of standard third-party delivery. DoorDash publishes marketplace rates of 15% to 30%. Run that spread across a $430 order and the difference stops being a rounding error.

Timing helped too. Catering orders arrive 48 hours or more ahead of service, which lets a kitchen prep against a known number instead of guessing.

Scale came with the deal. More than 140,000 restaurants on the platform. Gross bookings above $2.5 billion over the trailing twelve months, growing in the high teens. Roughly 900 employees, about 300 of them in Boston. Uber says the business is profitable, excluding certain expenses.

Dara Khosrowshahi, Chief Executive Officer of Uber, was direct about the appeal: "Catering is a big business, and can be a huge revenue stream for restaurants."

He is right, and that is the problem. Catering carried margin that delivery never did. One order feeds thirty people and repeats every Tuesday because somebody's office manager liked the sandwiches. Restaurants built those accounts slowly, one building at a time.

Nobody has said the commission will rise. Nobody has said it won't.

Demo Time, No Pressure

Take KitchenHub for a quick spin, zero commitment, just good vibes.

See Demo

Who Owns the Guest Record?

Read your agreement, because the answer is rarely the restaurant. Platforms collect order frequency, basket size, delivery address and payment method. Restaurants usually receive an aggregated or anonymized version of the same thing.

PYMNTS put it plainly in 2026: the platform knows who the customer is, and the restaurant often does not. For some fast-casual brands, third-party orders run above 50% of off-premise volume.

There is precedent for what happens once a platform decides your guest list is its asset. In March 2019, OpenTable changed its client agreement to stop restaurants passing diner data to other platforms without paying a fee. The target was SevenRooms, independent at the time. Steve Hafner, Chief Executive Officer of OpenTable, defended it on privacy grounds: "We can't stand by while third parties systemically harvest data from our systems without adequate user consent." Restaurant Dive covered the fight as it happened.

SevenRooms belongs to DoorDash now. Companies that fight over guest data have a way of ending up on the same balance sheet.

First-party data is the only version you control, and three questions tell you whether you have any. Can you export guest records with names, contact details and order history in a readable file? Does the agreement let you market to those guests directly? What happens to the records the day you leave?

Kevin Vasconi, chief digital and technology officer at Papa Johns, named the stake: "I think data is probably our greatest resource next to our people."

What Breaks First for a Virtual Brand?

Bargaining position. A virtual brand owns no channel outright, every customer arrives through somebody else's app, and the commission schedule is whatever the platform says it is.

A physical restaurant keeps a fallback. Guests walk in, call, book direct. A dining room is one channel nobody can buy out from under you.

Ghost kitchens carry the data problem in its purest form. The operator never meets a customer. Never takes a phone number, never hands anyone a receipt. Every record of every guest lives on a platform the operator does not own.

Picture a kitchen running four virtual concepts across two apps. If those apps also own the reservation layer and the catering layer, there is nowhere left to move volume when terms change. No "we'll push harder through catering instead."

How Do You Reduce Dependence on Delivery Marketplaces?

Treat marketplaces as customer acquisition, not as infrastructure. Four things stay portable no matter who buys whom: the phone number, the email list, the domain, and the record of who ordered what.

Start with what exports today. Pull your guest list from every system you run and check the format. A CSV with names, contact details, visit dates and order history travels anywhere. A dashboard view travels nowhere.

Read the exit terms before you need them. Who holds the domain your ordering page sits on. Whose developer account owns your branded app. What happens to gift card balances and loyalty points the day you leave. Those answers stop being negotiable after you sign.

Catering is the easiest channel to take direct, and most operators never try. Corporate clients order on a schedule and know exactly what they want. Pull the account list you already serve through a marketplace. Call the office managers. Give them a direct ordering link and a reason to use it. Commission-free catering platforms sell this exact move, and so does a decent page on your own domain.

Savneet Singh, Chief Executive Officer of PAR Technology, thinks the next shift in ordering tilts back toward operators: "That means the restaurant has the data, understands the customer's preferences, and doesn't pay a toll to another platform." Whether it does depends on who owns the channel when it arrives.

Uber's ezCater deal has not closed. Restaurants on that platform have a window, measured in months, to export their corporate client list while it still belongs to them.

Most will not use it.

The foodtech digest that gets read
Bi-weekly news and takes on what’s changing in restaurant industry.
I’m in