Every restaurant owner has had this month. Cheese is up, labor is up, the delivery platforms take their cut, and the menu prices are the same ones you set two years ago. You know you need to raise them. You’re also terrified that the day you do, the dining room empties out.
Here’s what actually happened when one pizzeria ran an across-the-board increase of roughly 4%, measured carefully over the following two months.
Week two looks like a disaster. It usually isn’t.
The first read after the increase was ugly. Orders down 8.3% year over year. If you’d stopped measuring there — and most owners do, because that’s when the panic peaks — you’d have rolled the whole thing back.
Six weeks later, orders had recovered to down 1.8%. By the following month, comparable-day traffic was actually flat to slightly up, with net sales up 13.6% year over year.
The sticker shock faded. It almost always does. What looks like a demand collapse in the first fortnight is usually your regulars noticing the new number, grumbling, and then continuing to come in. The short-run reaction and the long-run reaction are different animals, and the short-run one is the liar.
The practical rule: don’t judge a price increase for at least six weeks, and don’t judge it at all on a two-week window.
Measure it the right way or you’ll fool yourself
The same pizzeria’s raw month-over-month numbers said net sales were down 1.9%. The correctly measured number said they were up 1.6%. Same month, same data, opposite conclusion.
The difference was three things:
- Calendar alignment. June started on a Monday that year, giving the month an extra Tuesday and one fewer Sunday. Comparing calendar dates instead of weekday averages built a fake decline into the comparison.
- A structural change mixed in. Sunday lunch had been deliberately closed. That closure was the entire Sunday decline — nothing to do with price — but it dragged the whole month’s average down.
- Cherry-picked windows. An earlier internal report had claimed a huge profit win using a three-day stretch of cool weather. It was meaningless.
If you’re going to measure a price change, compare weekday-averaged year-over-year figures, exclude holidays that fell differently, and pull the same window on both sides. Otherwise you’re just reading noise and making decisions on it.
Check that your increase actually reached the register
This is the cheapest money in the building and almost nobody checks it.
When that pizzeria audited its point-of-sale system against the price changes it had intended to make, several never landed. The pepperoni slice was still at the old price while the cheese slice had gone up 12%. The large 14" pie, the cannoli, the cookies — all still ringing at old prices, weeks after the “increase.”
Pull your printed price change list. Ring up every item on it. Some percentage of your increase is probably sitting in a menu file nobody pushed live.
You may not be the cheap one
That pizzeria went into this convinced it was underpriced against the competition and needed a big catch-up increase. So the prices of every comparable local restaurant got pulled and put side by side.
The result contradicted the whole premise. On every shared pizza size, they were at the top of the local market — above every independent competitor, including the premium artisan shop across town, and level with the national chains’ list prices. The 16" cheese at $21.99 was three to five dollars above the nearest independent.
That changed the strategy completely. “We’re cheap, raise everything” became “we’re the price leader on purpose, and the job is making the premium visible” — ingredient story, menu language, marketing — rather than closing a gap that didn’t exist.
Before you decide how much room you have, go buy your competitors’ menus. Not what you remember them charging. What they charge today.
Watch the mix, but don’t misread it
After the increase, the pizzeria’s customers were clearly buying smaller pies. Large-pie share fell from 60% to 51%. The 18" and 16" were both down about 25% in units while the 12" was up 34%.
The obvious conclusion — the price increase drove people to smaller pizzas — turned out to be wrong. Pulling the full year showed large-pie share had been running in that lower band all year, well before the increase. The prior year’s 60% was the outlier, not the new number. It was a value-trading market shift, not a reaction to anything they’d done.
Two lessons. First, always check whether the “change” you’re seeing predates the thing you’re blaming. Second, look at the substitution math: large pies fell by 162 units while small pies rose by only 23. That’s not customers trading down, that’s whole-pie occasions disappearing entirely — a demand problem, not a pricing one, and the fix is completely different.
The counterintuitive part: your biggest item may be your worst
Run revenue per square inch on a pizza menu and something uncomfortable usually appears:
| Size | Revenue per square inch |
|---|---|
| 18" | $0.118 |
| 12" | $0.162 |
| 10" | $0.223 |
The giant pie everyone’s proud of is the margin laggard. The little one is the best deal in the building — for the restaurant.
This isn’t an argument for dropping your large sizes. It’s an argument for knowing which items actually carry the business, and for pricing your premium sizes steeply enough to be worth making. A wider gap between small and large simultaneously makes the small one look like the value option and fixes the economics on the big one.
Do the same exercise with plate cost on entrees, or per-ounce on drinks. The item you assume is the winner frequently isn’t.
What this adds up to
Inflation forces the increase. It doesn’t tell you how to do it. The operators who get burned are usually the ones who raise prices blind, panic at the two-week number, roll it back, and conclude “our customers won’t pay more” — when what actually happened is they never measured long enough to find out.
Raise deliberately. Verify it landed. Measure it honestly over a real window. Know where you actually sit in your market before assuming you have room, or assuming you don’t.
If you want help reading what your own numbers are saying, book a time.