Restauromics Weekly Update

On October 19 the National Restaurant Association rewrote its own math. The Census Bureau benchmarked its monthly restaurant sales survey to a new, more complete annual survey for the first time, and the numbers came down hard. The Association's 2026 sales projection dropped from $1.55 trillion to $1.36 trillion. That is roughly $190 billion, or about 12%, that never actually existed.

The revision reaches all the way back. Sales in 2021 were first reported at $1.06 trillion, a 24% jump over 2020. The revised number is $940 billion, an 11% gain. And the gap between the old and new figures widened every year after that. Translation: the post-pandemic recovery everyone celebrated was a lot slower than the headlines said.

The growth rate did not change. The Association still projects 4.3% nominal growth in 2026, which works out to just 0.8% once you take inflation out. Census also reported that eating and drinking place sales rose for an eighth straight month in September. So the industry is growing. It is just smaller, and growing slower, than the benchmarks you have been measuring yourself against.

Read the NRA revision analysis →

Olo CEO Noah Glass is launching a consumer app for delivery and takeout, and he is aiming it straight at the marketplaces. His pitch: aggregator commissions can hit 30% per order, so restaurants mark up their marketplace menus, so guests pay 15% to 30% more than they would ordering direct. Olo's app charges the guest a flat 5% network fee instead, shown at the top of the menu and at checkout.

About 20,000 locations across 250 brands already use it, out of roughly 90,000 Olo locations total. Orders land at the restaurant as if they came through its own channel, so the operator keeps the customer data. Delivery still runs through drivers from DoorDash, Uber Eats, and others via white-label services that charge a set fee per trip instead of a commission.

Glass wants the app to hit one billion orders a year by around 2030. He also admitted it is unlikely to change industry economics right away. For now it serves Olo's existing customers, which skews toward chains, with plans to open it to the wider industry later. No timeline yet.

Read about Olo's delivery app →

For two years the big burger chains tried to discount their way back to traffic. It did not work. "The deep value war is over, and all have lost," BTIG analyst Peter Saleh wrote this month. Chains got thinner margins and lower stock prices, not the crowds they paid for.

The retreat is everywhere. McDonald's pulled national advertising for its "$3 and Under" menu after about a third of franchisees did not run it as planned, and it spent its September investor day talking about food quality instead. Wendy's is backing away from Biggie Bags, which its own CEO called "value diluting." Domino's grew orders in Q3, but average tickets fell, and its CEO signaled fewer deals ahead.

Meanwhile, the chains that won did it with better food. Burger King posted record Q2 comps on the back of an improved Whopper, and Cava has priced below inflation since 2019 while its sales kept climbing. The new competitive weapon is quality, not coupons.

Read the value war analysis →

California's SB 68 already requires chains with 20 or more U.S. locations to tell guests in writing which of the nine major allergens are in every menu item served in the state. It is the first law of its kind. If you run one or two locations, it sounds like somebody else's problem. It is not, for long.

The FDA Food Code also requires operators to notify guests in writing about major allergens in unpackaged food. The Food Code is not law on its own, but most states adopt it into their health regulations, and the FDA's guidance says that notice should name the allergens in each specific item. The new 2026 Food Code, released September 17, keeps those requirements. Some states already go further.

Steritech food safety VP Chris Boyles, writing in Restaurant Dive on October 23, warns that operators and health inspectors may read the Food Code language differently, and he recommends getting a program in place within six to 12 months. The hard part is not the menu printout. It is keeping it accurate every time a supplier swaps a product or a cook tweaks a recipe.

Read the allergen labeling analysis →