Most loyalty platforms offer two mechanics and no guidance on choosing between them. The choice is not stylistic: pick the wrong one and the programme is either unfair to your best customers or trivially easy to game.
Which loyalty mechanic is right for my business?
The right mechanic is decided by how much your ticket size varies between visits. Low variance means stamps, high variance means points, and no relationship between spend and value means visits.
| Mechanic | Use when | Typical business | Fails when |
|---|---|---|---|
| Stamps | Every visit is worth about the same | Café, bakery, juice bar | A €2 espresso earns the same as a €40 order |
| Points | Ticket sizes vary a lot | Restaurant, retail, bar | Small tickets feel unrewarded, so occasional customers disengage |
| Visits | Attendance is the goal, not spend | Gym, salon, class studio | Customers can 'visit' without transacting |
When are stamps the right answer?
Stamps are right when a customer's purchases are interchangeable in value, because the mechanic quietly assumes every stamp is worth the same to you.
A café is the textbook case: drinks cluster within a euro or two of each other, so a free tenth drink is fair to everyone and legible at a glance. Legibility matters more than people expect, a customer who can see their progress without doing arithmetic stays engaged.
When should I use points instead?
Use points when a single transaction can be worth ten times another, because a stamp card in that setting either overpays small tickets or insults large ones.
The cost of points is comprehension. 'One point per euro, €10 off at 200 points' requires the customer to hold a conversion rate in their head. Keep the rate round, state the reward in currency rather than points, and never change the rate on existing customers.
When do visits make sense?
A stamp card makes sense when the behaviour you want is attendance and spend is either fixed or irrelevant, one stamp per visit, for a gym on a monthly membership, a class studio, a salon with a booking model.
The failure mode is enforcement: if a stamp can be recorded without a transaction, it will be. Tie the stamp to something that only happens when the customer is there.
Can I change mechanic later?
You can, but you should migrate rather than switch. Customers holding progress under the old mechanic must be converted at a rate that never reduces what they had, and told about it plainly. A programme that resets progress teaches customers that progress is not real, which is the one lesson a loyalty programme cannot survive.
How do I set the point value?
Use one point per unit of currency and set the reward at a round number, because the only property of a points scheme that matters operationally is whether a customer can estimate their progress without doing arithmetic.
One point per euro with a €10 reward at 200 points means the customer is earning a 5 per cent return, which is generous but legible. The same economics expressed as 7 points per euro and a reward at 1,340 points is incomprehensible, and a customer who cannot estimate their progress does not experience the goal-gradient effect that makes the mechanic work at all.
Two rules protect you afterwards. State the reward in currency on the card, "€10 off at 200 points" rather than in points alone. And never reduce the rate for customers who have already accumulated: if you must devalue, do it for new earnings only, and say so before it takes effect.
Can I combine mechanics?
Combining mechanics is technically possible on several platforms and almost always a mistake for an independent business, because every added rule has to be explained at a counter by a part-time staff member to a customer in a hurry.
The test is whether the whole programme fits in one spoken sentence. "Ten coffees, the eleventh is free" passes. "A point per euro, double on pastries before eleven, plus a stamp on any hot drink" does not, and the customer will simply stop tracking it.
If you need two behaviours rewarded, the cleaner answer is one mechanic plus occasional targeted offers through the card. The base scheme stays legible, and the promotion carries its own explanation each time it runs.
What does this look like in practice?
Three worked examples, using the variance test from the table above.
A café with €3 to €7 tickets
Variance is low, so stamps are correct. Ten stamps, the reward set to the customer's usual order rather than the cheapest item, and two stamps given at sign-up to trigger endowed progress. A regular visiting twice a week completes the card in about four weeks.
A restaurant with €15 to €120 tickets
Variance is high, so stamps would reward a solo coffee the same as a table of six. One point per euro with €10 off at 200 points is fair across the range, and lands for a typical couple after roughly three or four dinners.
A salon with six-week appointment cycles
Spend is fairly consistent but frequency is low, so a stamp card with one stamp per visit is the right unit and the threshold must be short. Six visits is about nine months, already at the limit of what anyone will pursue, so ten would be a mistake. Pair it with a wallet reminder before the next appointment is due, which is worth more than the reward itself in this category.
What about tiers?
Tiers work for businesses with thousands of customers and a wide spread of spending, and they are usually premature for an independent shop. A tier only means something if enough customers sit in each one to make the distinction visible.
A café with 200 loyalty members that introduces Bronze, Silver and Gold has created three labels and no behaviour change. The same effort spent on enrolling another hundred customers produces a measurably larger result.
If you do add a tier later, add exactly one, make it meaningful, and make the qualification obvious. A single "regulars" tier with a real benefit beats a four-level scheme nobody can locate themselves in.
How do I explain the mechanic to staff?
Write the programme as one sentence and give that sentence to staff verbatim, because whatever they improvise is what your customers will hear. The most carefully designed scheme is worth nothing if the person at the till describes it differently every shift.
Then answer the three questions customers reliably ask, in advance: what do I get, how long will it take, and what do you do with my number. Staff who can answer those without hesitating enrol substantially more customers than staff who have to find a manager.