How many stamps should a card have?
Ten is the default and it is often the wrong answer. The right number follows how often your customers already visit, not what the reward costs you.
The number of stamps on a card is usually picked by feel, and the feel is usually “ten, because that is what everyone does”. Ten is often right. It is right for a reason that has nothing to do with round numbers, and once you know the reason you can tell when it is wrong.
Key takeaways
- Size the card so a customer can finish it in roughly six to eight weeks at the rate they already visit. That is the whole rule.
- This means the number depends on visit frequency, not on what the reward costs. A three-times-a-week coffee regular needs around 20 stamps; a barber’s client needs three or four.
- Visible progress is what drives completion. In a field experiment, customers given a 10-space card with 2 stamps pre-filled completed it at 34% against 19% for an 8-space card needing the same eight purchases (Nunes and Drèze, Journal of Consumer Research, 2006).
- Ten stamps for a daily regular is not a loyalty program. It is a 10% discount handed to somebody who was already coming.
The rule underneath it
A card works when a customer can finish it before they forget about it.
That gives you an actual formula, and it uses a number you already know:
Stamps ≈ visits the customer already makes in six to eight weeks.
Six to eight weeks is roughly how long a goal stays live in someone’s head without a reminder. Longer than that and the card stops being a goal and starts being a piece of admin.
Worth saying plainly: that window is our working rule, not a measured finding. The research below supports the idea that visible progress drives completion. We have not found a study that puts a number on how long a reward stays motivating, so treat six to eight weeks as a starting point to test rather than a constant.
The failure at each end looks different, which is why both are easy to miss.
A card that is too long does not annoy anybody. It just quietly stops being a factor. The customer collects a few stamps, loses track of how far along they are, and behaves exactly as they would have with no card. You are running the program and paying for the software without buying any behavior change.
A card that is too short looks like a success. Redemptions are high, people mention it, and it feels popular. What is actually happening is that you are discounting visits that were going to happen anyway. The margin arithmetic is here.
Working it through
Take the visit rate first, then multiply.
| Trade | Typical visits | Sensible card |
|---|---|---|
| Coffee shop | 3× a week | 20–24 |
| Bakery | 1× a week | 6–8 |
| Takeaway | 2× a month | 4–6 |
| Barber | 1× every 4 weeks | 3–4 |
The coffee number surprises people. A daily regular hits ten stamps in under a fortnight, which means the reward arrives before they have changed any behavior. You have simply given a discount to someone who was already coming.
Use your own visit data rather than the table where you have it. These frequencies are illustrative starting points, not measured averages. The typical customer is the one to size for, not the most frequent one: a card built around your best customer will be far too long for everybody else.
The bottom two rows sit above what the rule alone would give, and that is deliberate. Applied strictly, a barber on a four-week cycle works out at under two stamps, which is not a card at all. Three or four is the practical floor: below that there is no progress for the customer to see, and the whole mechanism stops working. Where the arithmetic and the floor disagree, take the floor and accept a longer window.
Why progress matters more than the reward
The strongest evidence on card design is not about what the reward is. It is about whether people can see how far along they are.
In a field experiment at a professional car wash, Joseph Nunes and Xavier Drèze distributed 300 loyalty cards. One group received a card with eight blank spaces requiring eight purchases. The other received a card with ten spaces, two of them already stamped, so they also needed exactly eight purchases.
Identical effort, identical reward. Completion was 34% for the pre-filled card against 19% for the blank one.
The effect has a name: endowed progress. The practical reading for a shop is simpler than the psychology. A customer who can see where they are in a goal keeps going; one who cannot, drifts.
Two things follow for card length. A card short enough to show visible movement after a couple of visits holds attention. And starting somebody with a stamp already on the card, at the moment they join, costs you almost nothing and puts them past the hardest part.
When to go shorter than the rule
Three cases, and only three:
- A brand-new program. A short first card buys you proof that the thing works at all. Lengthen it on the second cycle.
- A trade with long gaps. A barber on a six-week cycle cannot be given a ten-stamp card; that is over a year, and nobody finishes it.
- A seasonal push. A deliberately short card with an end date is a different instrument, and it should look like one.
Outside those, the pull is nearly always toward making the card longer than instinct suggests, because instinct is anchored on ten.
What the reward should be
Not the cheapest thing you sell. The reward is the whole reason the card gets carried, and a customer who does the arithmetic on a 40p reward will stop.
The useful test: would you mention it out loud at the counter? “Tenth coffee free” passes. “5% off your eleventh visit” does not, and no sign will save it.
Two further constraints are worth applying. The reward should be a thing rather than a percentage, because a thing is easy to picture and a percentage needs working out. And it should cost you less than the margin on the visits the card adds, which is usually a lower bar than owners expect.
What happens when they finish
The common mistake is treating redemption as the end of the story.
A customer who has just redeemed is in your shop, pleased, and holding a card that now reads zero. That is the best moment in the entire cycle to start the next one, and the easiest thing in most programs to fix. If the card resets silently and nobody says anything, the momentum you spent eight weeks building is dropped on the spot.
There is research on exactly this pattern. Studying a real café reward program, Kivetz, Urminsky and Zheng found that members speed up as they approach a reward, then slow down immediately after earning one before re-accelerating toward the next. They also found that customers who accelerate more tend to stay and re-engage sooner. The post-reward dip is normal, in other words, and the practical response is to shorten it by starting the next card while the customer is still standing there.
It is also worth remembering what the card was for. The stamps were the reason somebody agreed to be recognized and messaged. The relationship does not end when the reward is handed over, and neither should the contact.
Frequently asked questions
Is ten stamps ever the right number? Often. Ten suits a customer who visits between one and two times a week, which covers a lot of bakeries, takeaways and casual restaurants. It is wrong specifically for the daily regular and for trades with month-long gaps.
What if my customers vary a lot in how often they come? Size for the typical customer rather than the extremes, and consider whether stamps are the right mechanic at all. Where spend and frequency both vary widely, points usually fit better. Stamps, points or vouchers sets out the choice.
Should I put a free stamp on the card when someone joins? Generally yes. It costs you a fraction of one reward, it gives a new member visible progress immediately, and the evidence above suggests visible progress is most of what drives completion.
Can I change the length after launching? Yes, and the second cycle is the natural moment. Avoid shortening a card people are partway through, because that reads as moving the goalposts even when it is in their favor.
Does a longer card mean fewer people finish? Up to a point that is the trade you are making, and it is the wrong thing to optimize. A card finished by everybody quickly is usually a card that changed nobody’s behavior. Judge it on repeat-visit rate rather than completion rate.
Where to go next
The length is one of four decisions. What a stamp card is actually for covers why you are running one at all, stamps, points or vouchers covers the mechanic, and loyalty program metrics covers how to tell whether it is working. The full sequence is in the complete guide.
Setting a card length, and changing it on the next cycle without reissuing anybody’s card, is what Passumo does. There are worked examples for cafés and for barbers.
Source: Joseph C. Nunes and Xavier Drèze, “The Endowed Progress Effect: How Artificial Advancement Increases Effort”, Journal of Consumer Research, Vol. 32, No. 4 (March 2006), pp. 504-512. Field experiment, 300 loyalty cards distributed at a professional car wash. Ran Kivetz, Oleg Urminsky and Yuhuang Zheng, “The Goal-Gradient Hypothesis Resurrected: Purchase Acceleration, Illusionary Goal Progress, and Customer Retention”, Journal of Marketing Research, Vol. 43, No. 1 (February 2006), pp. 39-58. Mixed method, including field data from a café “buy 10, get 1 free” program. Cited here for the direction of the effect only.