One of the first decisions you'll make when setting up a loyalty programme is your reward threshold, the number of stamps a customer needs to collect before they earn a freebie. It sounds simple, but getting it wrong can quietly undermine the whole system.
The psychology of progress
People are motivated by progress. A stamp card with six slots feels achievable. One with twelve feels like a marathon. The sweet spot is a threshold where customers can see the finish line from the start: close enough to feel attainable, far enough to be worth your while.
Research into what psychologists call the goal-gradient effect shows that people speed up as they approach a goal. A customer with five out of six stamps will visit sooner than one with two out of six. This is why giving a starter stamp, one free stamp just for signing up, is so effective. It puts the customer on the board immediately.
The numbers behind the decision
Let's say your average coffee costs three pounds fifty. If your reward is a free drink and your threshold is six stamps, you're giving away one drink for every six purchased, roughly a 14% discount. At eight stamps, it's closer to 11%. At ten, it's 9%.
That might sound like a small difference, but at scale it adds up. A shop serving 200 loyalty customers per week at a six-stamp threshold is giving away significantly more than one at ten stamps.
But here's the catch: a lower threshold means more frequent rewards, which means more frequent dopamine hits, which means stronger habit formation. The question isn't just "what can I afford?" It's "what drives the most return visits?"
The discount you quote is not the cost you pay
This is the piece that changes most people's answer, and it gets missed because the headline percentage is so easy to calculate.
That 14% is what the customer receives, measured at menu price. It is not what leaves your business. The free coffee costs you what it costs you to make, not what you sell it for. On a £3.50 drink at a 75% gross margin, the giveaway is about 88 pence of actual cost, set against the £21 that customer spent to earn it. That is nearer 4% of their spend than 14%.
Two things follow. The first is that a stamp card is a far cheaper instrument than it looks on paper, and shops routinely set thresholds too high because they are pricing the giveaway at retail. The second is the more important one: unlike a shelf-wide discount, this costs you nothing on the visits that never happened. A price cut pays out on every sale including the ones you would have made anyway. A completed card only ever pays out to somebody who came back six times.
What the arithmetic suggests
We have no completion data of our own to show you, and anyone who quotes you some without saying whose it is is worth asking. What we can do is reason from how often people actually visit, and that points at six to eight stamps for an independent coffee shop. Here is the reasoning:
Six stamps works well for shops with daily commuter traffic. These customers visit frequently, so they complete cards quickly and stay engaged. The faster reward cycle keeps the momentum going.
Eight stamps is the sweet spot for most independents. It's achievable within two to three weeks for a regular visitor, which keeps the goal visible. It also gives you a comfortable margin on the reward.
Ten or more stamps tends to cause drop-off. Customers lose interest or forget about the programme before they reach the finish line. Unless your average customer visits daily, ten stamps feels too far away.
Rough starting points by trade
The threshold that works follows the visit rhythm, not the price. A useful rule of thumb: pick the number that gets a typical regular to a reward in about two months.
- Daily or near-daily trades (coffee, bakeries, sandwich shops): eight to ten. A five-day-a-week customer finishes ten in a fortnight, so you can afford the longer card and still keep the finish in sight.
- Weekly or fortnightly (barbers, takeaways, pubs): six to eight. A four-week barber cycle makes even six stamps a six-month card, which is at the outer edge of what stays interesting.
- Monthly or longer (hair salons, nail bars, garages): five, sometimes four. Ten visits on a six-week cycle is well over a year, and nobody is pursuing a goal that far away.
Cross-check whatever you land on against the money using the stamp card calculator, but let the rhythm choose first. A threshold that is right on margin and wrong on timing produces a card nobody finishes, which costs you nothing in rewards and everything in point.
Don't forget the starter stamp
Whatever threshold you choose, consider giving new customers a free first stamp when they sign up. This does two things: it creates an immediate sense of progress, and it makes the remaining stamps feel more achievable. It is the best evidenced idea in this post, and it has a name: the endowed progress effect.
A card that shows "1 of 8" feels very different from one that shows "0 of 8." That one small gesture can meaningfully increase how many people complete their card.
You can always adjust
The beauty of a digital loyalty system is that you can change your threshold without printing new cards. If you start at eight stamps and find that completion rates are low, try dropping to six for a month and see what happens. Track the data, look at completion rates, and let the numbers guide you.
If you want to put your own numbers to it, the stamp card calculator shows what each threshold costs you and what it leaves behind.
Changing it later without annoying anyone
One practical wrinkle, since the whole appeal of digital is that you can change your mind.
If you lower the threshold, everybody currently holding a card gets closer to their reward, and that is straightforwardly good news you can tell them about. If you raise it, do not apply the new number to cards already in progress. A customer who was three stamps from a free coffee this morning and is now five away has been robbed, in the only sense that matters, which is how it feels. Let the existing cards finish on the old terms and start the new number with new customers. It costs you a handful of rewards and it is the difference between adjusting a scheme and breaking trust in it.
There's no universal right answer. It depends on your prices, your margins, and your customer behaviour. But with the right data, you can stop guessing and start optimising.