Restaurants Are Betting Big on AI. Only 5% Say It's Actually Working.

The restaurant industry's love affair with artificial intelligence has entered an uncomfortable phase: the morning after.

Qu's seventh annual State of Digital report, released March 19, surveyed 168 restaurant brands representing 94,000 fast-casual and QSR locations across the United States. The headline finding is one that should give every operator pause before signing their next AI vendor contract: 73% of brands are actively investing in AI or plan to start in 2026. But only 5% report measurable operational value or meaningful impact on the guest experience.

That is not a rounding error. It is a $62-billion-dollar industry pouring capital into technology that, by its own admission, has not yet delivered.

The Traffic Problem Driving the Spend#

The context for this AI rush is grim. According to the Qu benchmark, 57% of surveyed brands report a decline in guest traffic or visit frequency. Among QSR operators specifically, that figure climbs to 67%.

The National Restaurant Association's own data reinforces the picture. January 2026 marked the 12th consecutive month of net customer traffic declines across the industry. The NRA's Restaurant Performance Index stood at 99.3 in December, below the neutral 100 mark that separates expansion from contraction.

Operators are not investing in AI because they are flush with cash. They are investing because they are losing guests and need to find them again. Fifty-four percent of Qu's respondents cited food and commodity inflation as a headwind. Forty-five percent cited labor costs. Thirty-eight percent pointed to pressure on value perception.

"When restaurants face declining guest traffic, growth can't come from pricing alone," said Amir Hudda, CEO of Qu. "The guest experience must be improved across channels, from ordering to kitchen to fulfillment."

Where the Money Is Going#

The Qu data reveals a clear hierarchy in how restaurants are deploying AI budgets. Marketing, CRM, and personalization top the list at 53% of brands investing. Predictive analytics follows at 40%. Voice ordering sits at 39%, inventory and demand forecasting at 35%.

Kitchen automation lags at 23%, as does AI ordering agents at 23%. Computer vision for drive-thrus captures 19% of investment, while computer vision for kitchens draws 16%. Dynamic menu pricing, despite its theoretical promise, attracts only 13% of brands.

The pattern is telling. Restaurants are spending on AI that touches the customer before the order and the data after it, but investing far less in the operational core where food gets made and served. The 55% of brands that cite operational execution as their top barrier to better guest experience are, by and large, not directing their AI dollars at operations.

A separate survey from Informa Foodservice, conducted in partnership with PAR Technology and covering nearly 500 operators, tells the same story from a different angle. POS investment is now prioritized by 53% of operators, up from 40% last year. But digital ordering actually dropped from 38% to 29% on priority lists. Operators are pivoting from customer-facing flashiness to foundational infrastructure.

"You're seeing operators interested in quick ROI solutions like labor schedule optimization and buying the right amount of food at the right time," said Savneet Singh, CEO of PAR Technology.

The ROI Gap Nobody Wants to Talk About#

Here is the number that should be projected on the wall of every restaurant technology conference: among approximately 85 chains in the Qu study that are actively using AI, only 9% said it has had a meaningful or transformational impact. Thirty-three percent reported "emerging value." Forty-three percent reported "limited value."

Those figures do not describe a technology revolution. They describe a technology experiment.

The disconnect has a structural explanation. Thirty-seven percent of brands told Qu that fragmented systems and data prevent them from getting the most value out of their technology investments. Most restaurant chains work with a patchwork of vendors for POS, kitchen display, loyalty, inventory, and delivery. These systems often do not talk to each other. Layering AI on top of disconnected data is like building a navigation system without a map.

"Without that foundation, AI becomes another tool layered onto disconnected systems rather than a true growth engine," Hudda said.

Richard Del Valle, CIO of Bojangles, captured the operator perspective in Informa's report. He noted that while he will never be a fan of building proprietary technology from scratch, custom versions of off-the-shelf products represent a workable compromise for chains that need integration without the cost of ground-up development.

What Is Actually Working#

Despite the sobering ROI numbers, pockets of real progress exist.

Yum Brands has processed more than 2 million drive-thru orders through AI voice ordering across 300-plus Taco Bell locations. That is not a pilot. It is a deployment at meaningful scale, and it provides the kind of structured data that compounds in value over time.

Yum's China division has gone further, introducing Q-Smart, an AI assistant for restaurant managers that handles labor scheduling, inventory management, and food safety inspections through voice commands delivered via wireless earphones and smartwatches. The system targets the general manager's daily workflow rather than the guest-facing experience, a distinction that aligns with where the industry's actual bottlenecks live.

The NRA reports that 52% of operators using automation say they have seen faster service as a result. Self-service kiosks, a more mature technology than generative AI, continue to deliver measurable lifts: chains that have fully deployed kiosk ordering typically see 10% to 20% higher average check sizes compared to counter orders, driven by upsell prompts and reduced friction around add-ons.

The Spending Keeps Climbing#

None of the cautionary data is slowing down investment. Forty-eight percent of brands in the Qu study plan to increase technology spending in 2026. Among QSR brands specifically, that figure hits 54%.

Deloitte data cited by PYMNTS puts the number even higher: 80% of restaurant executives say they plan to increase AI spending in the next fiscal year. A Popmenu survey of 328 operators found 44% have already adopted AI, with another 25% intending to add it this year.

The investment is not irrational. Even at a 5% meaningful-impact rate, the operators who crack the code on AI-driven operations will hold a structural advantage in an industry where pre-tax profit margins average 4%, according to the NRA. In a business where food and labor each consume roughly 33 cents of every sales dollar, even marginal efficiency gains translate to outsized profit impact.

The risk is not that restaurants are investing in AI. It is that they are investing without the data infrastructure to make it work. Sixty-two percent of Qu respondents said improving order flow across all channels is their top priority for 2026. Fifty-two percent cited team workflow, training, and station efficiency. These are plumbing problems, not AI problems, and they need to be solved first.

The Operator Playbook#

For restaurant operators evaluating their own AI strategy, the Qu benchmark offers a practical framework.

First, fix the foundation. If your POS, kitchen display, and loyalty systems do not share data cleanly, no AI layer will compensate. The 37% of brands reporting fragmented systems as a barrier are telling you what not to do.

Second, prioritize back-of-house. The industry is overinvesting in marketing AI and underinvesting in operational AI. The 55% citing execution as their top barrier are not going to solve that problem with better personalization emails.

Third, set realistic timelines. The 43% of brands reporting "limited value" from AI are not necessarily making bad investments. They may be making early investments in technology that needs 18 to 24 months to show returns. The mistake is not the spend; it is expecting quarter-over-quarter ROI from systems that require behavioral change across thousands of crew members.

Fourth, watch the leaders. Yum Brands' 2-million-order voice AI dataset and Chipotle's robotic kitchen experiments are generating proprietary operational data that smaller chains cannot replicate. If you are a 50-unit brand, you do not need to be first. You need to be fast second, adopting proven technology after the megachains have absorbed the learning curve.

The restaurant technology market is projected to exceed $62 billion in 2026 for POS systems alone. The money is flowing. The question is whether the pipes are ready to carry it.

As Jen Kern, CMO of Qu, put it: "Hospitality wins. Brands that keep hospitality front and center while simplifying operations and thoughtfully integrating modern technology will lead the pack."

The data suggests most brands are still working on the simplifying part.