Restaurant loyalty beyond the discount: filling the empty table
Discounting rewards demand you already had. Here is what to reward instead, and how to move covers onto the nights that actually need them.
Most restaurant loyalty schemes are a discount with a card attached. Ten percent off, every visit, forever. It feels generous, it is easy to explain, and it is close to the worst possible use of the money.
The problem is not the size of the discount. It is when it lands. A standing percentage off pays out hardest on a Saturday at eight, when the room was full anyway, and does nothing at all on the Tuesday when it is half empty. You are buying covers you already had.
The useful question is narrower: what would make someone come on a night they would otherwise have stayed in?
Key takeaways
- A standing percentage discount pays out in proportion to demand, so it costs you most exactly when you need it least.
- Reward with food rather than money off. A £7.50 dessert costs you around £2.50 to make, so the customer receives three times what you spend.
- Restaurants are one of the few trades where points beat stamps, because a £20 cover and a £120 cover should not earn the same thing.
- Aim campaigns at a night, not at everyone. Filling Tuesday is worth more than thickening Saturday.
- Judge it on covers moved into quiet service, not on how many people joined.
What to reward instead
| Instead of | Give | Why it works |
|---|---|---|
| 10% off every bill | A dessert or aperitif at a set visit count | Costs you a third of its menu price |
| A discount on any night | The same reward, quiet nights only | Moves demand instead of subsidizing it |
| Money off the food | A chef’s course or a table upgrade | No cash comparison for the customer to make |
| A birthday discount | A birthday table with something on it | Books a party, not a solo saving |
The arithmetic that kills the percentage discount
Take a £35 average cover at roughly 70% gross margin: about £24.50 of the bill is yours before overheads.
A 10% standing discount hands back £3.50. Against gross margin that is not 10%, it is 14%, and you pay it on every cover including the ones that were never in doubt.
Now give a dessert instead. It lists at £7.50 and costs you about £2.50 in food. The customer receives something they would have paid £7.50 for; you spend £2.50.
The dessert is worth more to the customer and costs you less. That gap only exists because food carries a margin and cash does not.
Use points, not stamps
Most small shops should run stamps. Restaurants are the exception.
A stamp counts visits and treats them all alike, which is fine when everyone orders roughly the same thing. In a dining room one table spends £20 on two plates and another spends £120 on a tasting menu and wine. Rewarding both with one stamp overpays the first and insults the second.
Points track spend, which is what varies. The cost is legibility (a customer has to convert points into something real) so do two things:
- Quote the reward in food. “About four dinners to a free dessert course” beats “400 points”.
- Keep one rate. One point per pound, one reward threshold. The moment there are two rates, the scheme needs explaining at the pass.
The full comparison, including where stamps still win, is in stamps, points or vouchers.
Reward the night you need, not the night you have
This is the whole game, and it is the part a paper card cannot do.
A digital card knows who your regulars are and can carry an offer to them specifically. That turns loyalty from a standing cost into a scheduling tool:
- Quiet-night dining. A midweek reward that does not apply Friday or Saturday. Not a discount (the dessert, the aperitif, the extra course) but only Monday to Wednesday.
- New menu preview. When the menu changes, the people who already like your food are the cheapest possible test audience, and they tell other people. Send it to the top slice of your list a week before it goes public.
- VIP night. A closed or half-closed service for the fifty people who come most. It fills a dead night completely, costs you a normal service’s food, and it is the single most effective thing on this list.
- Win-back. Someone who came monthly and has not been in for ten weeks gets one reason to return. Nothing else on your list needs to hear it.
Each of these is a message to a segment, not a broadcast. A restaurant with 800 members that messages all 800 every time will be muted by most of them inside a month.
The channel follows the job. The new menu preview and the VIP invitation are email: they deserve more than one line, and a photograph of the dish does the selling. The midweek offer is a push to the card, because it is only true this week. The win-back is a text, because a table on a quiet Tuesday is worth more than the few pence a message costs, and a text is the one that gets read. Collect the number and the address on the join page so all three are available when you need them.
Getting people onto the card at the table
Restaurants have an advantage cafés do not: a seated customer with three idle minutes waiting for the bill.
That is the moment. A small code on the bill presenter, or on the table itself, with one line of copy that says what the reward is, not “join our loyalty program”, which promises admin. “Your fourth dinner comes with dessert” is the whole pitch.
Do not put the sign-up at the door. Nobody joins a scheme before they know whether they like the food.
The card goes straight into Apple Wallet or Google Wallet, so there is no app to download and no password to invent at the table, which is the step that loses most people. How the two wallets differ is covered in Apple Wallet vs Google Wallet.
What to watch
Sign-ups are not the number. Two are:
Covers on quiet services. Compare Monday-to-Wednesday covers against the same period before you started. This is what the program is for, and it is the only figure that justifies the reward spend.
Redemption rate. The share of earned rewards actually claimed. Very low means the threshold is too far away to feel reachable. Very high with no movement in quiet covers means you are giving desserts to people who were booking anyway. Tighten the offer to midweek.
Give it two full cycles. If your threshold is four dinners and your regulars come monthly, that is eight months before the numbers mean much. Set the threshold accordingly: it should be reachable inside a season, not a year.
Frequently asked questions
Does this work if most of my covers are one-off tourists? Partly. A destination restaurant with low repeat rates gets less from a card, so put the effort into the locals’ midweek trade and treat the weekend as a separate business.
Should the reward apply to drinks? An aperitif or a glass with the main works well and carries an even better margin than dessert. A blanket discount on the wine list does not: it is the percentage problem again, on your highest-value line.
What about bookings platforms and their own loyalty schemes? Those build loyalty to the platform, not to you. Your card is the thing that brings someone back without a commission attached, which is the same argument takeaways face with aggregators.
How many members before it is worth running campaigns? Around a hundred is enough to fill a quiet Tuesday if you aim at the right slice. Below that, concentrate on sign-ups and leave the messaging alone.
Where to go next
The mechanics (card type, thresholds, what a reward should cost) are in the complete guide to digital loyalty cards. The neighboring trades solve the same problem differently: pubs reward the regulars without discounting the bar, and takeaways use the card to pull orders back off the aggregators.
Passumo runs the card in Apple and Google Wallet with spend-based points, audience targeting for midweek campaigns and the counter materials in one place. There is a worked setup for restaurants.