What we guessed wrong, and right, about restaurant tech this year

We review our 2025 restaurant tech bingo card. | Photo: Shutterstock

Accountability is a core value here at Tech Check, and sometimes that means admitting when we are wrong. Earlier this year, we made five bold predictions about what would happen in restaurant technology in 2025. The good news is, we did not completely whiff, but we still plan to keep our day job.

Here’s a review of how we did, with the very minor caveat that things could technically still change in the five weeks between now and New Year’s.

Toast will buy something: Incorrect

I had guessed that Toast might use some of its cash flow to make an acquisition, particularly in the realm of enterprise restaurants as it looks to add more big-chain clients. There was speculation that it might buy Olo, but that didn’t happen. And ultimately, Toast has been able to keep growing its up-market business all by itself. Just a couple of weeks ago, it announced that TGI Fridays and the 100-unit Everbowl were signing up for its technology, joining a roster that also includes Applebee’s, Potbelly, Perkins and more.

Rather than acquisitions, Toast’s year was highlighted by a pair of high-profile partnerships: In August, it partnered with reservations platforms Resy and Tock with plans to leverage their customer data to help restaurants provide better service. And then earlier this month, it said it was joining forces with Uber to make it easier for restaurants to use the two services together, with additional functionality to come. 

A restaurant chain will hire a chief AI officer: Correct

The CAIO is a hot new job title in the age of AI, and it seemed only a matter of time before a restaurant chain got in on the action. None other than McDonald’s broke the seal in September, hiring Das Dasgupta as its first chief data, analytics and AI officer.

Dasgupta is coming off a brief stint as chief analytics officer for Starbucks and previously worked with McKinsey, Amazon and Viacom. Details on what his new job entails are slim: A LinkedIn post announcing his hire said he will work to grow the business “through data, analytics and AI,” which is literally just a repeat of the words in his title.

But we know McDonald’s is exploring multiple uses of AI, including computer vision, drive-thru voice bots and “virtual managers,” so it makes sense to have someone overseeing all of those efforts. 

Texas Roadhouse will join a delivery platform: Incorrect

On almost every Texas Roadhouse earnings call, an analyst asks if and when the steakhouse chain plans to start offering delivery. Here is how CEO Jerry Morgan answered that question the last time it came up, in August:

“I will continue to have conversations with operators. But as of right now, I think we’re holding the line on not doing delivery … unless there’s a real reason to do it. Individually, we will have some conversations. But as of right now, we have resisted going that route. We’re focused on providing our guests a great experience in the dining room and through our off-premise, through our pickup system and through the app and all of that. That’s where we’d really like to continue to focus as of right now.”

I had thought that growing competition in casual dining might push the segment leader to end this long-held stance in a bid to add revenue, like Olive Garden did last year. But Texas Roadhouse has kept chugging right along: Same-store sales rose 6.1% year over year last quarter. And it just does not like the idea of having someone else deliver its food.

I still think the chain gives in at some point, but I won’t pretend to have any idea when that might be. 

CloudKitchens will file for an IPO: Mostly incorrect

The Travis Kalanick-run ghost kitchen company did not file for an IPO this year. However, it is reportedly exploring the possibility of doing so with its Middle Eastern operations. As of September, it had several banks lined up, according to Bloomberg.

The ghost kitchen market is far more mature in countries like Saudi Arabia and the UAE, and CloudKitchens is partially backed by Saudi Arabia’s Private Investment Fund, so an IPO there is not all that surprising.

Here in the U.S., where delivery-only kitchens have largely been declared dead, CloudKitchens seems to still be growing: Its website now lists 93 locations all over the country, and there has been some real momentum in its adjacent office catering business, Picnic (which prompted Kalanick to post his first tweet in more than six years last week). 

Still, it’s not clear if a U.S. IPO is in the cards for CloudKitchens at this point.

First-party ordering will make strides: Mostly correct

In the first half of the year, this prediction was looking like a clear win. In April, I wrote in this column about how first-party ordering was gaining momentum, citing several data points showing that orders placed on a restaurant’s own website or app were growing (or expected to grow) faster than third-party channels.

However, as the year went on, the story began to look more complicated. 

We got some data showing that first-party ordering was losing one of its biggest expected advantages: lower prices. A study by Intouch Insight found that consumers were paying roughly the same for delivery orders from first- and third-party channels, and that they didn’t view one ordering method as better than the other.

Meanwhile, demand for DoorDash and Uber Eats appeared to accelerate. And we continued to hear some of the same complaints from restaurants about how hard it is to convert third-party delivery customers to first-party channels. Some operators have now given up entirely on converting people, and are instead treating third- and first-party customers as totally separate audiences.

Taken as a whole, I still think it’s fair to say first-party ordering made progress this year. Online ordering continues to grow, and restaurants are making major investments in loyalty programs. But it may have also taken a step back as delivery apps tighten their grip on consumers.

Final score? Let’s call it a 2 out of 5, give or take a few decimals. Stay tuned in the coming weeks for our 2026 predictions.