In this article
Every owner gets the pitch a few times a year: put a deal on our platform, we'll fill your quiet evenings, and you only pay when someone walks in. It sounds like free marketing. It's almost never free, and it's almost never marketing.
The problem isn't that a deal site lies. The problem is that you're shown just one number — the turnout, the full room, the thirty extra covers — and that's exactly the number that tells you nothing about whether it pays. The three numbers that decide that don't appear in the pitch at all.
The first is what actually lands after the discount AND the commission. A deal site stacks two things: a deep discount for the guest, and a commission for the platform on top of the already-reduced price. What reaches your account is often less than the plate cost to make.
The second is how many of those guests would have come anyway. Someone who buys a deal for a place they'd have visited regardless costs you no extra cover — they cost you the gap between a full price and a half one. And the third is how many deal guests ever return at full price. The whole promise turns on that, and it's precisely the number a deal site never guarantees.
This guide walks the seven numbers one by one, with the arithmetic attached. At the bottom you put your own deal into the calculator: your price, the discount, the commission, your food cost, and how many guests return. You'll see what you keep per deal cover, the return rate you'd need to break even, and what the promotion earns or costs you per year, all in. Everything runs in your own browser; nothing is sent or stored.
Why a deal cover looks free and isn't
Ask an owner after a deal promotion whether it paid, and the answer is usually: "well, the room was full." That's true, and it's exactly the trap. On the plate, a deal cover often roughly breaks even — you don't visibly lose on it, so it feels free. The money doesn't disappear at the table; it disappears in two places nobody adds up.
The first is cannibalisation: the guests who'd have come anyway. For them you swap a margin of about €24 for a deal margin of almost nothing. That gap, not the plate, is the real cost — and it appears on no receipt, because the guest simply pays with their coupon.
The second is the return that never comes. The whole promise of a deal site is that you meet new guests who then come back at full price. But a guest who hunts for a discount comes back for the next discount — yours or the neighbour's. The numbers below aren't against deals; they're against doing a deal without running it first.
The 7 numbers behind the deal
The first three explain where the money goes, the next two why the promise doesn't hold, and the last two what you do instead. At the bottom you run it with your own numbers.
1. −50%: the discount the guest sees
A deal site works because the discount is deep. Twenty per cent gets nobody off the sofa; fifty per cent does. A €45 menu listed at €22.50 is an easy yes for the buyer — and that's the whole point of the platform: it sells the feeling of a bargain, not an evening at your place.
For you, that deep discount means you've already given away half your selling price before anyone sits down. That can be a sensible investment — if you get something valuable back. The rest of this guide is about whether you do.
Remember this to start: the discount the guest sees is not the discount you carry. What you carry is the discount plus whatever the platform takes on top. That's the next number.
2. 40%: the commission the platform takes — on top of your discount
This is where almost every owner's mental maths goes wrong. You think: I give 50% off, so I keep half. But the platform charges its commission on the already-reduced price. On that €22.50, a deal site typically takes between 20 and 50% — call it 40%, and that's €9.00.
What reaches you then, from a €45 menu, is €13.50 including VAT. The discount and the commission together have eaten more than seventy per cent of your selling price before you've bought a single ingredient. And on that €13.50 you still owe VAT: net, about €11.16 is left.
Compare that with the commission on a booking platform, where you pay a few euros per booking on a guest who pays full price. A deal site is a different order of magnitude, because the commission comes on top of a halving.
3. ≈ €0: what's left once the plate is paid for
Now the sum that counts. From your €45 menu, €11.16 net reaches you. The plate itself — the purchase cost, excluding VAT — comes to about €13.02 at a 35% food cost. Which means that per deal cover you don't break even, you lose a little: the plate costs you more than what comes in.
That's exactly why a deal feels free and isn't. A loss of a euro or two per cover doesn't show up in a busy service; you see a full room, not an empty margin. The graphic below sets one cover side by side twice — at full price and on the deal — so you can see where the money goes.
And this is the kind case, because we only count the food cost. A deal evening that fills your room often needs extra staff too — and that loss isn't in here yet.
At the top, at full price; below, the same cover on a deal with 50% off and 40% commission. Same guest, same kitchen, two very different outcomes.
At full price, more than half is left. On the deal, the discount together with the commission is so large that what comes in doesn't even cover the plate — there's no margin left, there's one missing. And that's before the extra staff of a busy deal evening.
4. 35%: the guests who'd have come anyway
So far this has been about the guests you win. But some of the people who buy your coupon were coming regardless: a regular who sees the deal and thinks "why not", a local who was already considering it. Reckon on about a third.
For those guests you gain no extra cover — you swap a full margin for a deal margin. A guest who normally gave you €24 of margin now gives you almost nothing on the deal. That gap of roughly €26 per cover is the most expensive line of the whole promotion, and it's one you never see: the guest simply pays with their coupon and you don't notice.
This is exactly why a deal is most dangerous for a place that's already doing well. The more regulars you have, the greater the chance your coupon lands with someone who'd have paid your margin. You're paying a platform to give your own customers a discount.
5. 1 in 5: how many deal guests ever come back
The whole promise of a deal site is that the loss up front is made good by returns at the back: you meet new guests who then come back at full price. On paper that holds. In practice a minority ever return, and fewer still at full price — because someone who came for a discount is waiting for the next one.
The graphic below makes it concrete. Of a hundred deal covers, about thirty-five would have come anyway. Of the sixty-five genuinely new guests, about one in six comes back — ten or so. The rest came once, for the deal, and you never saw them again.
That handful is not nothing, and it's where the only real profit of a deal sits. But you paid a deep discount AND a commission on all hundred to meet those ten — and gave thirty-five regulars a discount along the way. Whether it adds up depends entirely on how well you hold on to those ten.
Who genuinely comes on top, and who you simply gave a discount to on a guest who was coming anyway.
You paid a deep discount AND a commission on all hundred to meet those green ten — and gave thirty-five regulars a discount they didn't need along the way. Whether the promotion pays depends entirely on how well you hold on to those ten at full price.
6. 3×: what filling your quiet hours yourself costs less
The real reason an owner reaches for a deal site is rarely "I want to give a discount". It's "my Tuesday and Wednesday are half empty and I don't know how to fill them". That's a fair problem — but a deal site is the most expensive way to solve it.
You can fill those same quiet hours yourself without putting a third party between you and your guest. A quiet-night menu on Tuesday, a fixed aperitif hour, a small gesture for anyone who books on a slow evening: it costs you the discount but not the commission — and the guest is immediately yours, with their details in your system rather than the platform's. The guide on filling quiet hours has nine ways that cost no commission.
The difference isn't only the commission. Someone who books through your own channel you can reach a second time; someone who comes via a deal site belongs to the platform. You pay to meet a guest you're then not allowed to contact. That's the same argument as a booking platform: own your bookings instead of renting them.
7. The exception: when a deal does pay
This isn't an argument that every deal is a loss. There's one case where it holds: you're new or unknown, a deal site genuinely reaches an audience you don't yet have, and you have a plan to keep those guests. Then the loss up front is a fair price for reach — not a marketing illusion but a calculated investment.
And if you do it, do it as customer acquisition, not as a discount. Cap the number of coupons so you don't give away your whole diary. Exclude your busy days and peak hours, so you're not handing the deal to guests who'd have come anyway. Build in a reason to return — a second visit on a soft condition, or simply an evening good enough to make the difference. And make sure you get the guest's details, not the platform, or you've bought an introduction you're not allowed to follow up.
Put those assumptions into the calculator in a moment and you'll see at once whether your case is the exception or the rule. The rule of thumb: a deal works not because the room was full, but because you knew who was in it and got them back a second time.
Run your own deal
Enter what a deal site would propose to you, and what you reckon about your guests. The fields are filled with a €45 menu at 50% off and 40% commission, so you can see how it reads — overwrite them with yours.
You get three numbers: what you keep per deal cover, the return rate you'd need to break even, and what the promotion earns or costs you per year all in — cannibalisation and returns included.
Deal true-cost calculator
Your own price, discount, commission and guests — and the difference in euros per year.
The deal
The guests
—
—
This counts only food cost as variable; extra staff on a busy deal evening isn't in it, so reality is usually a little heavier. Everything runs in your browser; nothing is sent or stored.
Two things to know as you read. The cover that "breaks even" isn't the problem — the problem is the guests who'd have come anyway, and the return that never comes. Set the share who'd have come to zero and your return rate high, and the number goes green: then it really is acquisition.
And watch the middle number. If the return rate you need is higher than what you realistically expect, the deal is a loss all in — however full the room was that evening. That one number decides it, not the turnout.
What to do with a deal proposal this week
Getting a pitch from a deal site? Decide not on the turnout they promise but on the arithmetic. This is the order.
Before you say yes — run it once
- Put the deal into the calculator above with the real commission the platform charges, not the discount they lead with.
- Estimate honestly what share of buyers are your regulars or locals — that number decides almost everything.
- Look at the return rate you'd need to break even. If it's higher than what you believe, the answer is no.
- Work out what the same discount would cost without commission through your own channel — usually that's the better deal.
If you do it anyway — as acquisition, not a discount
- Cap the number of coupons, so you don't give away your whole diary.
- Exclude your busy days and peak hours, so the deal goes to new guests and not to those who'd have come anyway.
- Make sure you get the guest's details and not the platform — otherwise you've bought an introduction you're not allowed to follow up.
- Build in a concrete reason to return, and measure after three months how many deal guests came back at full price.
Instead — fill your quiet hours yourself
- Make your own quiet-day offer and promote it through your own channels, so the guest is yours from the start.
- Ask every guest for their email or to book through your own system, so you can prompt a second visit yourself.
- Work out your customer value: one regular is worth more than ten deal guests, and that's where your marketing budget belongs.
- List the discounts you already give away — often you fill your quiet hours more cheaply by closing one existing leak than by running a deal.
The full room isn't the number that counts
Almost every owner who runs the numbers on a deal promotion finds the same thing: the cover roughly broke even, the room was full, and yet nothing was left at the bottom. Not because the deal went wrong, but because the turnout was the only thing measured, and turnout is precisely the number that proves nothing.
A deal site sells you a crowd that leaves the moment the coupon runs out. That can be worth it — once, as acquisition, with a plan to keep those guests. But as a standing fix for empty quiet hours it's the most expensive one there is: you pay a deep discount AND a commission to give your own customers a discount and bring in strangers who don't return.
Run your next proposal through the calculator above first. And then fill those same quiet hours yourself — with your own offer, your own guests and your own bookings. Someone who books through you, you can reach a second time; someone who comes via a platform belongs to the platform. That's the whole difference between a full evening and a guest who comes back.