How to Build a Restaurant Menu That Absorbs Rising Food Costs

Right-sized portions, a rotating value feature, and tiers that take the hit for you. Build the swing into the menu. Part four of four.

The short version: Build flexibility into the menu so it absorbs rising food costs without a full reprice. Offer a right-sized, smaller-protein portion. It carries a better margin, costs the guest less, doubles as a kids’ meal, and serves lighter and GLP-1 eaters. Run one rotating feature that always points at the cheapest ingredient that week, and use combos and good/better/best tiers so demand shifts to lower-cost options on its own when a protein spikes.

Last time was about moving your price with the clock, the fastest lever, because it’s just a decision. But it leaves you standing at the dial, reaching for it every time the market jumps. This week is the better version: a menu built so the swing has somewhere to go before a single crate ever spikes. You design it once, and it does the reacting for you.

Right-size the plate for the guest who now eats less

Here’s the one I’d build first. There’s a whole new guest walking in who eats noticeably less than they did two years ago. The GLP-1 wave is real and it isn’t slowing down. They don’t want the big platter. They’re a little relieved to pay for the right-sized one.

So give it to them. A garden bowl. A smaller portion of the expensive protein, built back up with the produce and grains that cost you a fraction of the meat. And here’s why it isn’t a discount: you’re cutting the single most expensive, most premium-laden input on the plate, your protein, so the small bowl can carry a better margin than the big one while costing the guest less. It earns more and asks less. That combination is rare. Take it when the market hands it to you.

Then it does a second job. The same bowl, the same prep, is your kids’ meal, and in a lot of our dining rooms the family table is the whole room. One item, two markets: the guest eating lighter and the child at the table. You priced once and sold twice.

Let one corner of the menu chase the market so the rest doesn’t have to

You don’t have to reprice the whole menu every time a crate of something spikes. You need one corner that’s allowed to move, so the core can hold steady.

Run a rotating feature that always points at whatever’s cheapest this week: the fruit at peak season, the cheese that’s gone soft, the value protein. The board moves. Your anchors don’t. Guests get a fresh reason to come back, and you get permission to follow the market without ever touching the prices people have memorized.

Two more valves while we’re here. The combo is a shock absorber: people will price a single item to the penny, but almost nobody prices out a bundle, so the combo is where you can move a little when costs move, and it’s where the cheap, high-margin drink rides along to soak up the swing. And tiering builds the valve right into the menu. Think a small, a standard, and a loaded, or a value protein, a mid, and a premium. When beef spikes, traffic slides to the lower tier on its own. When produce is cheap, the loaded version suddenly looks like a steal. You’re not reacting to the market. You built a menu that already did.

None of this is reacting. It’s deciding ahead of time

Every move here is the same move the whole series has been making: refusing to let the menu run on whatever it happened to inherit. A frozen menu hands the entire price swing to you to eat. A menu that moves hands it to structure you designed on purpose: windows, sizes, tiers, a feature that chases value so the core never has to.

You’ll never get it perfect, and that was the point back in part one: a menu set in January is a guess about a year you haven’t lived yet. The operators who come out ahead don’t price right once. They build a menu that can move when the year does, and then they keep reading it.

The market won’t hold still for you. Your menu doesn’t have to either.

Pete Deserto is the founder of Urban Food AI, built for halal operators. urbanfoodai.com