Berry AI ads

I’ve been keeping a running list of QSR AI stuff that makes me stop and go, okay, that’s actually pretty cool. The list got long enough that I figured I’d share it. There’s voice AI, robots at the fry station, some pretty incredible deployments at scale. Here’s what I’ve been watching in Q3 ↓

Prime Day used to be where you waited to buy the TV. Now? Protein shakes, hydration packets and everyday essentials are topping the list. Groceries and household essentials now account for 1 in 3 units sold on Amazon in the U.S. People are shopping differently, too. They’re spending less, comparing more and waiting for the right deal. And that’s where this gets interesting for QSR. This is an industry that has always been really good at the impulse buy. You’re hungry, you see the sign, you pull in. On the retail side, Amazon has spent decades teaching us a very different way to shop, and that’s having lasting effects. So as more everyday purchases (including food) move into that world, we wanted to know if some of those habits are following us to the drive-thru.

Drive through Drive-offs. Say that 5x fast. Apologies… we had to lighten the mood, especially after we did the math on these. The average drive-thru loses 5.7 cars a day to customers who see the line and decide it's not worth the wait. At an average order cost of $15, an operator can be staring down the barrel of a $31,000 a year per restaurant loss. Ouch. And those customers were right there. Hungry, at your restaurant, ready to order. Then they weren't. That’s a lot of almost-revenue, and it adds up differently for every operation. So we took a closer look at what it could mean for operators and built a free calculator to run the numbers for your own stores. Full breakdown ↓

Nine months. That’s roughly how long Starbucks’ inventory-scanning AI lasted after rolling out nationwide. Maybe inventory counting is simply one of the harder vision AI use cases to scale. Maybe the pilot didn’t account for enough real-world variation. Or maybe it just needed a little more time to brew.☕ We took a closer look, and there are a few things QSR operators should pay attention to.

Asian/noodle concepts grew sales 7.8% in 2025, making them the fastest-growing major limited-service category. And the rest of the growth leaderboard is pretty interesting too. Coffee/café grew 6.1%. Chicken grew 5.3%. Mexican-inspired concepts grew 4.7%. That's a pretty varied group of restaurants to find near the top of the same list. So we started looking at what they have in common: how customers define value, how much control they have over their order, how easily the menus can change and, naturally, what people are saying on Reddit. Let’s be clear here: the classics are still the classics. There's just a lot more room at the table now.

Most operators can tell you their average service time. Very few can tell you what thirty seconds of it is worth across the estate. Three inputs, no form, under two minutes