A loyalty reward that feels generous to the customer and safe to your P&L usually lands at an effective discount of 10% to 15%, funded by giving away your highest-margin item (a drink, not food), on a stamp count that matches your customer's realistic visit frequency. The trick most cafe owners miss: the reward pricing barely matters if enrolment friction kills signups. A programme with a 12% effective discount and 40% signup rate beats a 9% programme that needs an app download and converts at 8%, every single time.
- Aim for a 10 to 15% effective discount. Any less feels stingy; any more starts eating margin you cannot afford.
- Reward drinks, not food. Espresso-based drinks carry 80 to 90% gross margin; a filled croissant is closer to 35 to 45%.
- Match stamp count to real visit frequency. Ten stamps for a twice-a-week regular is fine; ten stamps for a once-a-fortnight visitor is a dead card.
- Signup friction is a cost. If joining needs an app download, expect 8 to 15% conversion. If it saves straight to Apple or Google Wallet, expect 30 to 50%.
- Nudge redemptions to slow periods so the giveaway lands when the barista is not already buried.
What your free coffee actually costs#
Start with the honest maths. A £3.50 flat white has a COGS of roughly 40 to 60p once you count beans, milk, cup, lid and a slice of the barista's minute. Call it 50p. Give one away for every nine bought, and you have spent 50p to earn £31.50 of revenue. That is a 1.6% cost of sales on rewarded transactions, or a 10% effective discount from the customer's point of view (they got their tenth drink free).
The customer's perception and your P&L live in two different columns. That gap is the entire reason loyalty works in cafes and does not work in, say, jewellery.
| Reward | Menu price | Approx COGS | Gross margin given away | Perceived generosity |
|---|---|---|---|---|
| Free flat white (after 9) | £3.50 | £0.50 | £0.50 | High |
| Free filter coffee (after 9) | £2.80 | £0.30 | £0.30 | Medium |
| Free filled croissant (after 9) | £4.20 | £2.50 | £2.50 | High |
| 10% off next order | varies | varies | 10% of every rewarded bill | Low |
| Free pastry with any drink | £2.80 | £1.60 | £1.60 | Medium |
Notice the free croissant. Customers rate it as generous as the free coffee, but you are giving away five times the COGS. If food is under 30% of your revenue, rewarding drinks is a no-brainer. If you are a food-led cafe where a filled sandwich makes up half of tickets, that changes: a free drink with any food purchase can actually pull food sales up while giving away the cheap thing.
Stamps or points: which structure protects margin better#
Stamps are simpler and cap your exposure. Ten stamps, one reward, done. Points scale with spend, which sounds fairer but exposes you: a customer buying a £15 catering round earns more toward their reward than someone buying a £3 espresso, even though the espresso drinker is a more valuable regular. For most independents, stamps win. If you have a wide price range (drinks, sandwiches, retail bags of beans), points may reflect real value better. There is more nuance in choosing between stamps and points for your cafe.
| Factor | Stamp card | Points |
|---|---|---|
| Customer understanding | Instant | Requires learning the ratio |
| Margin exposure | Fixed and predictable | Scales with spend |
| Best for | Drink-led cafes with a tight price range | Food-led or mixed venues with wider tickets |
| Feels generous? | Yes, visual progress is powerful | Only if ratio is clear |
| Peak-hour cost risk | Lower | Higher (bigger baskets earn more) |
How many stamps? Match the card to the customer, not the other way round#
The default is ten. It is also often wrong. Think about your actual visitor patterns:
- Twice-a-week regular: a 10-stamp card gets redeemed every 5 weeks. Motivating.
- Once-a-week regular: a 10-stamp card gets redeemed every 10 weeks. Borderline.
- Fortnightly visitor: a 10-stamp card takes 20 weeks. Dead card. They will forget it exists.
- Daily commuter: a 10-stamp card redeems every 2 weeks. Too easy, over-rewards.
For most independent cafes, an 8-stamp card is the sweet spot. It cuts time-to-reward by 20% versus ten, which sharpens the sense of progress, and the extra 2.5% effective discount is a fair price for higher retention.
"The best reward structure is the one your customers actually complete. A generous programme they abandon is worse than a modest programme they finish."
The cost competitors don't mention: peak-hour redemptions#
A free coffee at 8:12am on a Tuesday is more expensive than a free coffee at 3:30pm. Not in COGS: in queue time, kitchen chaos and the paying customer behind who watched a barista stop to scan a phone. Peak-hour redemption is a real operational tax and almost no loyalty guide addresses it in plain language.
You have two levers:
- Nudge earning toward slow periods. "Double stamps Monday to Friday before 11am" pulls loyal customers into your quiet window. The reward is still redeemable any time, so it does not feel restrictive.
- Reward the slow menu. If your afternoon filter coffee is under-selling, make it the free reward. Peak-time customers will still redeem for their usual, but a nudge toward the off-peak drink shifts some redemptions where you want them.
The maths only works if customers actually join#
This is where most cafe loyalty guides go quiet. You can design the most elegant reward structure in the world; if the signup step is "download our app, create an account, verify email", you will lose 85% of interested customers at the queue. StampMe's own analysis flags enrolment as the single biggest determinant of whether a programme returns its cost.
Run the numbers. Two cafes, both serve 200 daily customers, both offer a free drink after 9:
| Cafe A (app required) | Cafe B (wallet, no app) | |
|---|---|---|
| Signup conversion at counter | 8% | 40% |
| Monthly new members (200/day) | ~480 | ~2,400 |
| Effective discount | 10% | 12% |
| Members active after 90 days | ~30% | ~55% |
| Net repeat-visit lift after 6 months | Marginal | Meaningful |
Cafe B gives away a slightly bigger discount and wins on every metric that matters, because five times more people actually joined. A programme that lives in Apple Wallet or Google Wallet, saved from a QR code at the till in under 15 seconds, is a different product from one that asks for an app store trip. The reward pricing debate is downstream of that decision.
A one-page calculator for your reward#
- Pick the reward item. Aim for 80%+ gross margin. For most cafes that means an espresso-based drink.
- Work out its COGS. Beans, milk, cup, lid, syrup if relevant. Be honest.
- Divide COGS by (menu price × stamp count minus one). That is your effective cost of sales on rewarded transactions.
- Sense-check the effective discount. 1 ÷ stamp count = the customer's perceived discount. Target 10 to 15%.
- Check the redemption window. Estimate how many weeks a typical customer takes to complete the card. Under 6 weeks: motivating. Over 12: rethink.
Worked example, entirely hypothetical: an 8-stamp card, free £3.50 flat white as reward, 50p COGS. Effective discount to customer: 12.5%. Cost of sales on rewarded transactions: about 2%. A twice-weekly customer completes the card in 4 weeks. That is a healthy programme on paper. Whether it works in your shop depends on how many people join, which depends on how much friction stands between them and their wallet.
What is a fair reward for an independent cafe loyalty programme?
A free drink after 8 to 10 paid drinks is the standard, producing an effective discount of 10 to 12.5%. For most cafes this costs 1 to 2% of rewarded revenue in COGS, well within margin tolerance.
Should I reward food or drinks?
Drinks, unless food is over 30% of your revenue. Espresso-based drinks carry 80 to 90% gross margin, so giving one away is the cheapest way to feel generous. Food rewards can work but should be tied to a drink purchase to protect ticket size.
How do I stop customers redeeming during my busiest hour?
You cannot stop it, but you can nudge earning behaviour toward quiet periods (double stamps before 11am, for example) so heavy users build their reward on off-peak visits. Blocking peak-hour redemption feels stingy and rarely helps.
Is a paper stamp card still fine?
Fine to run, hard to measure. You will not know your redemption rate, active member count or repeat-visit lift. A digital card in Apple or Google Wallet costs about the same to run and tells you whether the programme is working.
How long should it take a customer to earn their reward?
Between 4 and 8 weeks for a typical regular. Faster than that and you are over-rewarding. Slower than that and most customers will disengage before they finish the card.