In this article
An all-you-can-eat buffet doesn't sell a dish, it sells a promise: eat as much as you want for one fixed price. That promise is exactly why it's the one line on your menu where you don't know what you sold until the guest gets up.
On a regular menu you know your cost before the order leaves the kitchen: one plate, one price, one margin. On a buffet you set a price before you know what the guest will eat — and that one price has to cover the guest who goes back twice the same as the guest who goes back six times. That's not a food-cost problem, it's a risk-pooling problem: you're bundling risk across a whole room of guests, and your price has to cover the tail of that distribution, not the middle.
The thread running through all of it is something hospitality rarely says out loud: a buffet doesn't attract the average eater by accident. Whoever can eat unlimited food for one fixed price is precisely the guest who values that most — and that's a different mechanism than 'some people just eat more'. It's the same logic an insurer relies on: whoever picks the most generous plan is, on average, also the one most likely to use it.
This guide walks through seven numbers that together explain why a buffet price that looks fine on paper still leaks margin in practice — from the guest who breaks your price to the table that sits three-quarters of an hour longer than it would on a regular menu. At the bottom you plug your own buffet into the calculator — your guest mix, your chafing-dish loss, your VAT and your covers — and get the price you actually need, in money per year.
Everything runs in your own browser: nothing is sent anywhere and nothing is stored. The numbers below are guide ranges for an independently run European venue — your concept, your city and your suppliers are the final judge.
Why 'cost per guest' is already the wrong question
A portion on a regular menu has one cost: the kitchen sets the quantity, you set the price, and both are fixed before service starts. On a buffet the guest sets the quantity — so 'cost per guest' isn't a cost at all, it's an average across a group of people who have nothing in common except paying the same price.
That's why you split your guests into three groups: light eaters, average eaters and heavy eaters. Not because that neatly summarises reality, but because one averaged number can look healthy while a single group alone makes your buffet unprofitable — and the average will never tell you that.
Every group below gets its own target zone: your target food cost, plus and minus a margin, because your guest mix shifts from one service to the next. Inside that zone a group is healthy. Above it, a group is eating your price — and that's what this guide is about.
The ultimate guide Menu Engineering & Beverages: 6 Steps to More Margin From menu layout to pour cost: everything your menu can earn you, in one guide. Open the guideThe 7 numbers, and what each one costs you
They run in the order they stack up: from how a buffet attracts its guests, to the table that ends up sitting three-quarters of an hour longer than it would on a regular menu.
1. Your buffet doesn't price a plate — it prices a distribution
Set your buffet at €27.30 and your target food cost at 33%, and the arithmetic looks simple: cost divided by revenue, done. But that cost is not a fixed number — it depends entirely on who is sitting at the table. A light eater (roughly 25% of your guests) eats about €4.40 of ingredients. An average eater (half your guests) eats about €6.95. A heavy eater — the last quarter — eats about €11.35. Add the loss at the chafing dish (see number 5) and those three amounts become €5.06, €7.99 and €13.05.
On the exact same €27.30 ticket, that gives three completely different food costs: 21.0% for the light eater, 33.2% for the average eater, and 54.3% for the heavy eater — more than twenty points above your 33% target. The figure below sets the three side by side, against your target zone of 29–37%.
At €27.30 a ticket and a 33% target food cost (dashed line), each guest produces a different food cost — the hatched section is what a guest eats above your target.
All three pay the exact same €27.30. The light eater hands you almost 12 points of margin for free; the heavy eater alone eats more than twenty points above your target — and none of that shows up in your average food cost.
2. A buffet doesn't get the average eater by accident — it gets adverse selection
Whoever can eat unlimited food for one fixed price isn't a random guest who happens to be hungry that night. It's the guest who already knew they'd eat a lot, and chose the buffet precisely because of it, instead of a menu where every extra portion costs extra. That's the same mechanism an insurer relies on: whoever picks the most generous plan is, on average, also the one most likely to use it — the pricing structure itself attracts exactly the risk it can least afford.
That means your '25% heavy eaters' isn't a random slice of the population — it's a group your own concept concentrates. A steakhouse buffet pulls heavier eaters than a lunch buffet built around soup and salad, and a buffet known for its seafood tower pulls them harder still.
So the share you type into the calculator below isn't a textbook number — it's something you have to measure at your own buffet. It shifts by concept, by city and by day of the week.
3. Once someone has paid, they eat until it 'feels worth it' — not until they're full
Just and Wansink's 2011 study on flat-rate pricing at pizza buffets (Review of Economics and Statistics) found a striking pattern: guests who paid the full price upfront ate more than guests who got the same meal at a discount — but rated the meal lower afterwards. The feeling that the price has to be 'worth it' outweighs the feeling of being full.
That effect isn't greed, it's a predictable consequence of how you pay. Charge by weight or by plate and every extra bite costs money again, which naturally slows people down. Charge one fixed amount upfront and that brake disappears entirely — the sunk cost is already spent, so eating more feels like winning rather than spending.
One countermeasure is cheaper than it sounds: smaller plates and cutlery at the buffet measurably shrink the portions people serve themselves, without anyone feeling restricted — the same effect plate-size research has long documented for self-service in general.
4. Heavy eaters don't load up evenly — they hit your most expensive line first
The extra cost of a heavy eater isn't spread evenly across your whole buffet. Whoever can serve themselves unlimited food for one fixed price fills their plate with the most expensive items first — the meat, the fish, the seafood tower — and only fills up on bread and salad afterwards, if there's room left. The marginal portion on your most expensive line is exactly the portion that gets repeated most often.
The classic counter-move is about sequence, not restriction: put the cheap, filling items — bread, soup, salad, potatoes — at the front of the line, and the most expensive proteins at the back. A guest who has already served themselves bread and salad twice usually reaches for a third helping of lobster less often.
The numbers in the calculator below already assume this typical skew — your heavy-eater group costs more precisely because it eats disproportionately from your expensive line. Without a deliberate line sequence, that skew only gets worse.
5. A buffet loses more food to the pass than a plated menu ever does
The whole promise of a buffet is that it always looks full. A chafing dish sitting empty for five minutes reads to a guest as a buffet that's failing, so every kitchen over-produces as insurance against exactly that moment. That surplus is product you bought and paid for, and nobody ever puts it on a plate.
This page uses 15% as a working figure for that loss — the sum of over-production, portions that dry out or cool down and get swapped, and whatever is still sitting in the trays at close. Buffet service structurally generates more unsold food than a plated menu where every dish is made to order.
The figure below breaks down where one €27.30 ticket actually goes, for each of the three guests: VAT first, then the ingredient, then the loss at the pass, and whatever's left is your margin.
VAT first, then the ingredient, then what you lose at the pass — whatever's left is your margin, per guest.
For the heavy eater, there's almost nothing left after VAT, ingredient and loss — and that's before you even count the longer table time from number 6.
6. The table sits longer — and that's a second bill nobody writes down
Even with a perfectly healthy food cost, a buffet isn't automatically as profitable as a regular menu, because the clock keeps running. A buffet table sits longer on average than a table on a regular menu — guests walk back and forth, chat between rounds, and there's no service pacing the meal.
In the example on this page, a buffet table sits for 110 minutes on average, against 75 minutes for a comparable table on a regular menu with an average check of €34 ex VAT. That gives roughly €13.12 of revenue per seat-hour for the buffet, against €27.20 for the à la carte table — roughly half. Anyone who wants the full formula behind that number will find it in the guide on RevPASH.
You don't fix that gap by changing the buffet itself — a guest who lingers for three-quarters of an hour often does so precisely because the buffet is enjoyable — but by how many buffet services you schedule alongside your regular menu each week.
7. The price has to cover the tail, not the middle — here's the formula
Everything above comes together in one calculation: take each group's cost per guest including loss, weight it by that group's share, and divide the result by your target food cost. That gives you the revenue you actually need; add VAT back and you have the price that belongs on your menu.
In the example on this page, that produces a required price of about €29.33 against the current €27.30 — an increase of barely 7%, even though the heavy-eater group runs more than twenty points above target. That's the whole point of this guide: the fix is rarely 'charge a lot more', because the blended average already looked healthy. The real work is knowing which lever is actually broken — price, mix, loss or table time — before you touch any of them.
Plug in your own numbers below — your guest mix, your loss, your VAT, your covers and your table times — and get the price, the food cost per group, and the difference in money per year.
Plug your own buffet into the calculator
Fill in what your buffet actually looks like: your guest mix, what each group eats in ingredients, your loss at the pass, your VAT, your target food cost and your covers. The three groups start filled in with the worked example above — overwrite them with your own counts.
You'll get the food cost per group against your own zone, your blended food cost, the price you actually need, what your heavy eaters cost you per year above target, and revenue per seat-hour for your buffet against a regular menu.
Buffet pricing scan
Three guest groups, your own VAT and covers, and the difference in money per year.
Fill in your ticket price, VAT, loss and target first — that sets the zone every guest group below is measured against.
—
The zone is a guide range for an independently run European venue and doesn't account for your concept, your city or your supplier terms. Everything runs in your browser; nothing is sent or stored.
Two things worth keeping apart. What your heavy eaters cost is a mix problem: that group simply eats more than your price planned for, and you fix it with portion control, line sequence or a separate price point — not a higher price for everyone. What your price is short by is a price problem: even with a perfect mix, your ticket doesn't hit your target.
And revenue per seat-hour is a third, separate story: even with a healthy food cost, a table that sits three-quarters of an hour longer earns less per hour than the same table would on a regular menu. You don't fix that with the buffet itself, but with how many buffet services you run each week.
What to do this week, this month and this quarter
Nobody fixes seven numbers at once. This order works, because every step makes the next one measurable.
This week — measure your real mix
- Count, on one busy service, how many times each guest goes back to the buffet — that's your first real guest mix, not a guess.
- Put your current ticket price, VAT and target food cost into the calculator above and see which group falls outside its zone.
- Look at your line sequence: are the cheap, filling dishes ahead of your most expensive line?
- Note how much is still sitting in the chafing dishes at the end of service — that's your first loss number.
This month — measure instead of guessing
- Weigh or count, across three services, what gets thrown out at close, and fill in your real loss percentage.
- Compare your table time on a buffet night with a regular menu night — a simple clock at order time is enough.
- Recalculate your heavy-eater group with the real numbers and see what it costs you per year above target.
- Test one change to your line sequence or portioning, and measure the same service again a month later.
This quarter — lock it in, then price
- Lay the past months' measurements side by side and look at the trend, not one service.
- Fix your mix and loss problems first, before touching your price — a price increase doesn't fix a portion problem.
- Only now adjust your ticket price, and only if the calculator points to a real price problem.
- Put the scan into your quarterly rhythm alongside your prime cost — a buffet belongs in the same conversation as the rest of your menu.
The margin isn't in your ticket price — it's in your guest mix
Almost every owner who runs this calculation finds the same pattern: the average food cost looks healthy, and the margin still disappears. Not because the price is wrong, but because one group of guests — usually a quarter of the room — eats a price that was calculated for everyone at once.
That's good news, because the problem is rarely where you'd look for it. A different line sequence, smaller plates at the buffet, and actually counting once a month instead of guessing — that's the whole fix, and it works before you ever have to touch your price.
Do the same with the rest of your menu next: the choice between prix fixe and à la carte and your tasting menu are the same question in different clothes: which menu format lets you decide what a guest eats, instead of the other way around?