Most loyalty programs fail quietly. They are not badly built. They are built with a reward too far away, launched without telling anyone, and abandoned three months later because "it did not seem to do much".
This is the order that avoids that.
Step 1: decide what you actually want it to do
Write down one sentence before you touch any software.
Almost every small business wants one of these:
- more visits from the customers you already have
- bigger baskets from the customers you already have
- fewer customers drifting away after two or three visits
These need different designs. If you want frequency, count visits. If you want basket size, count spend. If you want to stop the drift, put the first reward very close to the start so a new customer hits it before they forget about you.
Trying to do all three produces a program that does none of them well.
Step 2: choose stamps or points
This is the only structural decision that really matters at the start.
Stamps count visits. One stamp per qualifying purchase, a free item when the card fills. Choose this if most of your orders are close in value.
Points count money. Points per pound, a reward at a threshold. Choose this if your bill swings widely.
A quick test: look at your last fifty transactions. If the largest is less than about three times the smallest, use stamps. If it is more, use points, or use stamps restricted to one category.
That last option is underused. A bakery can stamp hot drinks only. A garden centre can stamp coffee and plants separately. You get the simplicity of stamps without paying a full reward to someone buying a single roll.
The other structures, tiers and paid memberships, are worth knowing about but are rarely right for a first program. Loyalty program examples covers all six with worked cases.
Step 3: set the reward and the threshold
Two numbers, and they are linked.
Start from what you can afford to give back. For most hospitality and services, somewhere between 5 and 12 percent of a member's spend is sustainable. Below that, customers do not notice. Above it, you are buying visits that would have happened anyway.
Then work backwards to a threshold that a regular can reach in six to eight weeks.
Say a customer visits twice a week. A ten stamp card takes five weeks. Good. Say a barber's customer visits monthly. A ten stamp card takes ten months, which is not a loyalty program, it is a filing cabinet. Drop it to six.
Run the calculation for your actual visit frequency rather than copying someone else's card.
The two stamp head start
If you can, start members with a stamp or two already on the card, and raise the threshold to match. Twelve stamps with two given free beats ten stamps from zero, even though the customer buys the same number of coffees either way.
This is not a trick, or not only a trick. It is a well documented effect: people are more motivated to finish something they have already started. Nunes and Drèze showed it with car wash cards in 2006 and it has held up since.
Make the first reward reachable
If you can, put a smaller reward partway. Something at four stamps on a twelve stamp card. The customer learns that the program is real and that you honour it, before they have invested two months in it.
Step 4: decide the rules before a customer asks
Write these down. Staff will be asked all of them.
- One stamp per visit, or one per item?
- Does a £2 purchase earn the same as a £20 one?
- Can a customer stamp twice in one day?
- Do stamps expire? After how long?
- Can two people combine cards?
- What happens if someone loses their card?
The one that causes the most friction is same day repeat stamping. A customer buying four coffees for the office reasonably expects four stamps. A customer coming back three times in a day is gaming it. Pick a rule, usually one stamp per transaction with a daily cap, and tell staff so they are not deciding it individually under pressure.
On expiry, be careful. Expiring stamps protects you from an endless liability but it is the single most common reason customers get annoyed. If you set one, make it generous, at least a year of inactivity, and tell people before it happens.
Step 5: pick how customers carry it
Three options, and the trade is the same each time: cost against knowing anything.
Paper cards cost pennies and need no setup. They also tell you nothing. You cannot see who your regulars are, you cannot contact anyone, and a meaningful share of cards get lost or go through a wash cycle. They are also trivially forged.
An app puts everything on the customer's phone, but asks them to download something. For a local business that ask is heavy. Most people will not install an app for one coffee shop.
A browser or wallet based card sits between the two. The customer scans a QR code and joins in the browser without installing anything, and can add the card to Apple Wallet or Google Wallet if they want it on the lock screen. This is what StampClub does, and it is the shape most independents end up at, because the signup friction is closest to paper while the data is closest to an app.
Whatever you choose, time the signup. If joining takes more than about thirty seconds at a busy counter, staff will stop offering it, and you will conclude that customers were not interested.
Step 6: give staff one sentence
This step gets skipped and it is the one that determines whether any of the rest matters.
Do not brief staff on the program. Give them the exact words.
"Are you collecting stamps? It's ten for a free coffee, takes about ten seconds."
That is it. One sentence, said at the same moment every time, usually while the card machine is thinking.
Two things help it stick. Put the QR code where the customer already looks, next to the card reader rather than on a wall behind the till. And check the signup rate weekly for the first month, because a rate that plateaus early is almost always a staff prompting problem rather than a customer interest problem.
We wrote about the training side in getting staff to promote loyalty naturally.
Step 7: launch to your regulars first
Do not open with a poster campaign.
Spend the first two weeks signing up the people already walking through the door. They join at a much higher rate than strangers, they will actually redeem, and they give you a real signup rate to measure against before you add any external promotion.
You also want the first few redemptions to happen while you are watching. The redemption moment is where operational problems show up: staff unsure whether to accept, a customer who thinks they had more stamps, a reward that turns out to be awkward to give away during a rush.
Step 8: check four numbers after two months
Not before. The first cohort needs time to fill a card at their natural rate.
Signup rate. What share of transactions involve a member. Under about 25 percent after two months means it is not being offered.
Visits per member, before and after joining. This is the number that tells you whether the program works. Everything else is context.
Redemption rate. The share of completed cards that get claimed. Low redemption is not free money. It means people lost interest before the finish line, which usually means the threshold is too high.
Margin given away. Total reward cost divided by total member spend. This should land near the percentage you chose in step 3. If it is much higher, your rules are leaking somewhere, often through same day stamping.
What to change if it is not working
Low signups: it is the ask, not the offer. Change what staff say and where the QR code sits before you touch the reward.
High signups, low redemptions: the threshold is too far away. Lower it, or add an interim reward partway.
Good redemptions, no change in visit frequency: you are rewarding people who were already coming every week. Move the reward toward behaviour you want more of, such as a bonus on your quietest day, rather than raising the value.
Everything fine for three months, then a slow decline: this is normal and it is a communication problem. Members forget. A single message to people who have not visited in six weeks usually recovers a useful share of them.
A worked example
A cafe with an average order of £4.20 and a coffee that costs them £1.30 to make.
Ten stamps, tenth drink free. Reward cost per completed card: £1.30 against roughly £37.80 of member spend, which is about 3.4 percent. Comfortably inside a sensible range, and there is room to be more generous.
So they go to nine stamps and start members with one already filled. Effectively eight paid drinks per reward, about 3.8 percent given away, and the card feels noticeably faster to the customer.
A twice weekly regular finishes in four weeks. That is the number that makes it work.
Where to go next
If you have not settled on a structure yet, start with loyalty program examples. If you want the cost side in more detail before committing, read how much a loyalty program costs. If you are moving off a paper card you already run, the paper to digital checklist covers honouring stamps customers already hold.
And when you are ready to build one, creating a loyalty program with StampClub takes about ten minutes.