how many stamps should a coffee loyalty card need?

Enough that the free drink is comfortably covered by the paid visits before it, usually somewhere between eight and ten for a standard coffee. Set it against your true marginal cost, not the menu price, and give a head-start stamp on joining so the goal feels reachable from day one.

is a paper stamp card or an app better for a small cafe?

Paper is fine for a single site where you know your regulars and will not use visit data anyway. Go digital only if you will actually act on what it tells you, like nudging lapsed customers or filling quiet hours. Do not take on app admin for the sake of looking modern.

will a loyalty program just cut my margins?

It can, if it rewards visits you already had. The regular who comes daily was not going anywhere, so their free drinks are margin given away. It pays off only when it lifts frequency or moves people into quiet hours by more than the rewards cost. Measure that, do not assume it.

should the reward be a free drink or a discount on every purchase?

A free drink after several visits, in almost every case. It feels like a genuine treat while costing you only one drink's marginal cost across many paid visits. A standing percentage off comes straight off every transaction, including full-price ones, and reads as a price cut rather than a reward.

how do i use loyalty to fill my quiet afternoons?

Attach the extra earning to the slow windows. Double stamps between two and four, a bonus for a Monday visit, or an afternoon pastry-and-coffee reward. You move existing customers into hours your staff and oven are already paid for, which is close to pure margin, instead of discounting the peak you never had to buy.

how do i know if my loyalty program is actually working?

Look at frequency and quiet-hour trade, not sign-ups or redemptions. The question is whether your existing regulars now come more often and whether dead patches filled in. If visits are flat while free drinks go out, the scheme is a cost, not a growth tool, and needs reworking or retiring.

Playbook

What loyalty programs actually work for a coffee shop?

October 30, 2026 · 8 minute read

The short answer

The scheme that works buys you one more visit a week from people who already like you: a simple buy-so-many, get-one-free reward that feels generous but costs you little, pointed at your quiet hours. It should not be a standing discount that trains every customer to wait for a freebie, or that trims the margin on regulars who would have paid full price anyway.

What you are actually buying

A coffee shop runs on repeat custom. Someone who likes your flat white does not need persuading, they need a small nudge to come three times a week instead of two. That is the whole job of a loyalty scheme. Not new faces, not a viral moment, just a little more frequency from people already walking past your door.

The reason this beats a discount push is the arithmetic of retention. The Bain and Company work associated with Fred Reichheld found that lifting customer retention by five per cent can raise profit by more than a quarter, because a regular costs almost nothing to serve and buys predictably. A loyalty scheme is a retention tool wearing a marketing hat. Judge it on whether your existing regulars come more often, not on sign-up numbers.

The right target for a cafe scheme

One extra visit a week from an existing regular, not a first visit from a stranger
Frequency and habit, not headline discount depth
Your quieter hours, not the queue that already forms at 8am
People who like the product, not bargain hunters passing through

The maths of a reward that feels generous but costs little

The trick every good coffee scheme uses is the gap between what a free drink is worth to the customer and what it costs you. Take an illustrative flat white at 3.50 pounds. The milk, beans and cup behind it cost you well under a pound. A buy-nine-get-the-tenth-free card hands the customer 3.50 pounds of perceived value, but your actual outlay is only the marginal cost of that one free drink, spread across ten paid visits. Your real discount is small, single digits as a percentage of their spend, while it reads as a proper treat.

Contrast that with a flat ten per cent off every cup. That comes straight off every transaction, including the ones people would happily have paid full price for, and it never feels like a reward. It just feels like your price dropped. The free-drink model concentrates the giveaway into a moment that feels like a gift, and only after the customer has already given you nine visits.

Run the numbers before you print anything

Know the true marginal cost of your reward drink, not its menu price
Set the threshold so the free item lands after enough paid visits to cover it easily
Exclude or cap high-cost items so a large oat mocha does not become your standing giveaway
Compare the free-item model against a flat discount on real transaction data, not gut feel

Give them a head start

How you draw the card changes how many people finish it. In a 2006 study in the Journal of Consumer Research, Joseph Nunes and Xavier Dreze ran two car wash loyalty cards. One needed eight stamps from scratch. The other needed ten but arrived with two already stamped, so both demanded the same eight purchases. The pre-stamped card was completed by about a third of customers against under a fifth for the blank one, and those customers came back faster.

The lesson for a cafe is cheap to apply. A card that shows some progress already made pulls harder than an empty grid, even when the work required is identical. Start people partway, or hand a stamp on the first visit for joining. It costs you nothing and it makes the goal feel close enough to bother chasing.

Paper versus digital, and when each makes sense

Paper cards are cheap, instant and need no sign-up. They also sit forgotten in coat pockets, get lost, and tell you nothing. You cannot see who is close to a reward, who has drifted away, or what your best regulars actually buy. For a single site with a tight counter and a chatty owner who recognises faces, paper is often enough, and its honesty is a feature.

Digital changes the trade. A phone-based or wallet card is harder to lose and, crucially, it captures visit patterns, so you can message the regular who has not been in for a fortnight, or push a quiet-Tuesday offer to people who only ever come at weekends. The cost is real, though. Someone has to run it, the customer has to bother installing or scanning it, and you take on data responsibilities. Do not adopt an app because it sounds modern. Adopt it only if you will actually use the data it gives you.

Choosing between them

Paper suits a single site, low volume, and an owner who knows the regulars by name
Digital earns its keep only if you will act on the visit data, not just collect it
Wallet passes on the phone avoid a separate app download, which is the main drop-off point
Whatever you pick, it must survive a Saturday rush without slowing the till

Point it at your quiet hours

A reward aimed at the moment you are already slammed is wasted. If there is a queue out the door at half eight, you do not need to incentivise those visits, they are happening anyway, and a stamp there just discounts demand you already own. The value sits in your dead patches: the mid-afternoon lull, the slow Monday, the gap before the after-school trade.

Tie the extra earning to those windows. A double stamp between two and four, a bonus for a Tuesday visit, a pastry-and-coffee reward that only bites in the afternoon. You are moving existing customers into hours when your staff and your oven are already paid for, which is close to pure margin. This is also where a scheme quietly protects you from the worst discounting mistake, rewarding the peak you never had to buy.

The honest costs and the regulars trap

Two things go wrong often enough to name plainly. The first is admin. Every scheme is a small ongoing job: replacing lost cards, settling disputes at the till, chasing an app that will not scan, checking the sums still work. If nobody owns that job it rots, and a half-run loyalty scheme is worse than none, because it annoys the people you most wanted to keep.

The second is subtler. A loyalty scheme can quietly hand money to customers who were never going anywhere. Your daily regular was going to buy that coffee with or without a stamp, so every reward they earn is margin you gave away for a visit you already had. That is fine if the scheme also lifts frequency or fills quiet hours by enough to cover it. It is a slow leak if it does not. And never train the whole room to wait for the freebie by making the reward the reason to come rather than a thank you for coming. The product is the reason. The stamp is the nudge.

Signs your scheme is leaking, not earning

Redemptions are up but total visits and quiet-hour trade are flat
Regulars talk about the reward more than the coffee
Nobody clearly owns the admin, so cards and disputes pile up at the till
You cannot say, on data, whether frequency actually rose after launch

Common questions

how many stamps should a coffee loyalty card need?: Enough that the free drink is comfortably covered by the paid visits before it, usually somewhere between eight and ten for a standard coffee. Set it against your true marginal cost, not the menu price, and give a head-start stamp on joining so the goal feels reachable from day one.
is a paper stamp card or an app better for a small cafe?: Paper is fine for a single site where you know your regulars and will not use visit data anyway. Go digital only if you will actually act on what it tells you, like nudging lapsed customers or filling quiet hours. Do not take on app admin for the sake of looking modern.
will a loyalty program just cut my margins?: It can, if it rewards visits you already had. The regular who comes daily was not going anywhere, so their free drinks are margin given away. It pays off only when it lifts frequency or moves people into quiet hours by more than the rewards cost. Measure that, do not assume it.
should the reward be a free drink or a discount on every purchase?: A free drink after several visits, in almost every case. It feels like a genuine treat while costing you only one drink's marginal cost across many paid visits. A standing percentage off comes straight off every transaction, including full-price ones, and reads as a price cut rather than a reward.
how do i use loyalty to fill my quiet afternoons?: Attach the extra earning to the slow windows. Double stamps between two and four, a bonus for a Monday visit, or an afternoon pastry-and-coffee reward. You move existing customers into hours your staff and oven are already paid for, which is close to pure margin, instead of discounting the peak you never had to buy.
how do i know if my loyalty program is actually working?: Look at frequency and quiet-hour trade, not sign-ups or redemptions. The question is whether your existing regulars now come more often and whether dead patches filled in. If visits are flat while free drinks go out, the scheme is a cost, not a growth tool, and needs reworking or retiring.