Restaurant Pricing by Daypart: The Same Dish, Two Different Economics

Restaurant economics are usually counted per dish: cost of ingredients against menu price, a food cost percentage, a decision. It is a reasonable starting point and it misses the single largest variable in the business — when the dish is sold.

The ingredients cost the same at three in the afternoon as they do at eight in the evening. Everything else does not. Rent, the chef, the waiters and the dishwasher are paid for the hour whether the room is full or empty, and those costs land on whatever number of guests happens to walk in.

Split the day before you touch the menu

Most restaurants have three or four distinct dayparts, and they behave like separate businesses: a slow morning, a busy lunch, a dead afternoon, a profitable evening. Averaging them into one monthly food cost hides the fact that one of them is subsidising the rest.

Start by splitting revenue and covers by daypart for a normal week. Then take your fixed costs — rent, salaried kitchen staff, utilities, administrative overheads — and distribute them by the hours each daypart occupies, not by its share of revenue. Distributing by revenue is circular: it makes the busy hours look expensive and the empty ones look cheap, which is precisely backwards.

The same dish, two different results

A pasta dish sells for 320 and costs 96 in ingredients — a 30% food cost, which looks healthy. The kitchen and floor cost 2,800 an hour to run, regardless of guests.

At eight in the evening the restaurant serves 40 covers an hour. The hourly labour and overhead spread across them comes to 70 per guest. The dish leaves 320 − 96 − 70 = 154.

At three in the afternoon the same restaurant serves 6 covers an hour. The same 2,800 now spreads to 467 per guest. The dish leaves 320 − 96 − 467 = −243. Every plate sold in that hour makes the day worse, and no food cost percentage will ever reveal it.

What dynamic pricing actually fixes

Dynamic pricing in restaurants has a bad reputation because it is often read as «charge more when people are desperate». Done from the numbers above, it is the opposite — it is a way to make an empty hour cost you less.

The realistic moves are three. Fill the empty hour with a cheaper, faster offer that still clears the variable cost: a set lunch that leaves 90 per guest is far better than an empty chair that leaves nothing. Shorten the empty hour — if the afternoon cannot be filled, it may be cheaper to close the kitchen and cut the shift. Price the peak properly, because the evening is where the margin has to be earned and a menu priced for the quiet hours gives it away.

Do not confuse a discount with a loss

The set lunch works only if it clears the variable cost and contributes something to the fixed one. Before discounting, separate the two: ingredients, the extra waiter you call in, the packaging — these are variable and must be covered. Rent and the salaried chef are paid anyway, so any contribution above the variable cost improves the day, even if the dish looks «unprofitable» at full cost.

This is the same distinction that decides whether a low-margin order is worth taking at all — the logic is set out in counting prime cost after every shift.

Watch it by shift, not by month

A monthly P&L shows that the restaurant earned or did not. It cannot show that Tuesday afternoons have been losing money since spring. By the time a bad daypart is visible in the monthly numbers it has been running for weeks.

Counting revenue, covers and labour by shift turns this into something you notice within days. It also settles arguments about staffing: a shift that is genuinely unprofitable is an argument with numbers, not with the chef — see scheduling to demand.

Where to start

Take one normal week. Split it into your real dayparts and write down, for each: revenue, covers, and the hours of kitchen and floor labour it consumed. Divide fixed costs by hours, not by revenue. You will usually find one daypart that carries the restaurant and one that quietly eats the margin.

In Finmap you see revenue and costs by day and by shift, so a loss-making hour shows up while you can still change the schedule. Try it free for 7 days.