Guide · 5 min
How many stamps before the first reward?
Work it out from what the reward costs you to make, not from what the cafe down the road does. Below about six stamps most programmes stop paying for themselves. Above about twelve, people stop believing they will ever get there.
Most businesses pick a stamp threshold by looking at what the venue down the road does. That number encodes somebody else’s margins, somebody else’s visit frequency, and quite possibly somebody else’s guess.
Here is a way to arrive at your own.
Start from what the reward actually costs you
The reward is not free to you, but it costs far less than its menu price. Say you sell a flat white for $5.50. The cup, milk, beans and labour come to $1.65. That $1.65 is what the free one costs you.
Work out what the visits earned
At a threshold of ten, a customer spends roughly ten times their average ticket to reach the reward. If that average is $6, they have generated $60 of revenue and you have given back $1.65 of cost.
That is 2.75 percent. Compare it against what you would pay for the same outcome through advertising, and it is difficult to lose that comparison.
The two boundaries
The arithmetic keeps working as the threshold falls, but two things happen at the edges that arithmetic does not capture.
Below about six stamps, the reward stops feeling earned. It reads as a discount, and discounts train customers to wait for discounts. You also start giving value to people who would have come anyway, which is the fastest way to make a loyalty programme unprofitable.
Above about twelve, the finish line stops feeling reachable. A customer who visits weekly is looking at three months of commitment for one free coffee, and most of them will quietly stop counting. A programme nobody completes is a liability sitting on your conscience and, in some accounting treatments, on your books.
Use your own visit frequency as the check
Work out how many stamps a regular collects in a normal week. Then see how many weeks it takes them to finish.
Ten stamps at two a week is five weeks, which is fine. Ten stamps at one a fortnight is five months, which is not. If it comes to more than about six weeks, either lower the number needed or give a stamp per item, so a bigger order earns more than one.
A worked example
A lunch bar with a $14 average ticket and 60 percent gross margin on a $12 rewarded item. The reward costs $4.80. At eight stamps the customer has spent $112, so the programme costs 4.3 percent of that revenue. Regulars come twice a week, so eight stamps is four weeks. That is a programme that works.
Move it to fifteen stamps and it costs you less on paper. In practice it is worse. It is now two months of visits, and the number of people finishing will tell you so.
Common questions
Is nine stamps standard?
It is common, which is not the same thing. It became common in cafes because the maths happens to work at typical coffee margins, not because it is a rule.
Should the reward be the same item?
Usually yes. A free version of the thing they already buy is easy to understand and its cost to you is known precisely.
What if nobody is finishing a card?
Either your threshold is too high for your visit frequency, or people are not being reminded they have a card at all. Check completion rates before you cut the threshold.