How to Test a Menu Price Increase Before You Roll It Out

Why a test beats an opinion

Price conversations in restaurants are usually settled by the loudest anecdote: a regular who complained, a server who thinks the portion is already small, an owner who remembers a price rise that went badly years ago. None of that is data about your current guests.

A test gives you three things an opinion cannot:

It also limits the blast radius. One dish at a slightly wrong price for two weeks is a small experiment; a whole-menu increase that goes wrong is a season of lost contribution.

• A number for unit sales before and after the change, on the same dish and the same dayparts.
• A contribution figure, which is the only number that decides whether the change was worth it.
• A rollback rule that is triggered by evidence instead of by nerves.

Step 1: pick one dish and one new price

Choose an item that meets four conditions:

Then choose the new price. A practical first test is a 3–5% move on the menu price, rounded to the menu convention you already use. Do not change the portion, the garnish or the recipe at the same time, or you will not know what the guests reacted to.

• It sells steadily, ideally 150 or more units a week. Low-volume items produce noisy results.
• It has volume in more than one daypart, or you accept testing it in a single daypart only.
• It has a documented plate cost, so contribution per sale is known rather than guessed.
• It is not the item guests mention by name on review sites. Start where a small change will not become a story.

Step 2: write the thresholds before you start

This is the step people skip, and it is the one that makes the test useful. Before the first sale at the new price, write down:

The maximum unit loss formula is:

new break-even units = old units × old contribution ÷ new contribution

Illustrative example. A pasta dish sells 210 units a week at $24.00 with $12.40 contribution. The price rises to $25.00, so the new contribution is $13.40 if food cost is unchanged.

• Baseline units per week, from at least four weeks of data.
• Baseline contribution per unit, from the current plate cost and price.
• Maximum acceptable unit loss, calculated from the contribution change.
• The measurement window, typically two weeks after a one-week transition.
• Pass, fail and ambiguous rules, including what happens if the result is in the middle.
• The rollback trigger, usually a unit loss larger than the maximum for two consecutive weeks.
• Break-even units: 210 × 12.40 ÷ 13.40 = 194.3, so about 194 units.
• Maximum unit loss: 210 − 194 = 16 units a week, or 7.6%.
• Pass: units hold at 195 or more, contribution above $2,600 a week.
• Fail: units fall below 194 for two consecutive weeks.
• Ambiguous: units between 194 and 200 with contribution roughly flat — extend one week before deciding.

Step 3: run the test without contaminating it

Mechanics matter as much as the math:

• Change the POS price and the printed menu in the same service. A mismatch between the check and the menu destroys trust faster than any price rise.
• Brief the team first, including the reason and the exact wording for a guest question: "Yes, that came into effect this week." No apologies on the floor.
• Keep the portion and plating identical. If the plate changes, the test is measuring something else.
• Avoid event weeks, holidays and promotion periods. Compare like with like.
• Check the mix, not just the units. If total pasta sales fall but guests trade to a higher-contribution dish, the menu is doing better than the unit count suggests.

Step 4: read the result correctly

After two weeks, compare four numbers with baseline: units, contribution per unit, total contribution for the dish, and the dish's share of total covers. Then apply the rules you wrote.

A pass is not only about this dish. A clean test tells you whether this menu category has room, and it gives you a template for the next item.

• Pass: roll the price into the next menu print and keep monitoring for four weeks.
• Fail: roll back to the original price, and note what the guests told you.
• Ambiguous: extend one week, or revert if the operations cost of the test is getting high.

Step 5: keep the test honest

Three habits keep price tests from becoming wishful thinking:

When a test passes, take the corrected contribution into the rest of the menu work. The menu pricing strategies guide covers the wider structure, and the profit margin calculator helps you see what the change does at the whole-menu level rather than on one line.

• Write the thresholds down first. Recomputing them after the fact is how a fail becomes a pass.
• Give it enough time but not too much. Two weeks beyond a one-week transition is usually enough for regulars to notice. Longer windows let seasonality into the result.
• Run one test at a time. Two simultaneous price changes make both results unreadable.

Limitations and assumptions

• A single-dish test cannot tell you what would happen across the whole menu. Guest tolerance, competitor prices and traffic patterns differ by item and by market.
• Contribution per unit assumes the plate cost is current. If the recipe cost is stale, the threshold is wrong before the test starts.
• Small samples are unreliable. A dish selling 60 units a week needs a longer window or a bigger effect to produce a trustworthy answer.
• Regulars adjust over time, and the first two weeks after a price increase are not the same as month three. Closing the test is not the same as ending the review.
• This protocol does not cover menu engineering decisions such as removing an item, changing its portion or repositioning it on the menu. Those deserve their own test.