Almost no loyalty programme fails in the software. It fails in week two, when the person who set it up stops mentioning it while taking payment. This guide is ordered by when things happen, not by the order of the menu.
Day 0
How many stamps and what you give away. The rule that works: the card should be finishable in about two months at the rate your people really return. Work out the cost of the prize, not its menu price, and divide it by the stamps: that is what each visit costs you.
The usual mistake: Setting twelve stamps “to make it worth it”. At one visit a month that is twelve months, and almost nobody gets there: the card is abandoned by the fourth.
Done when: I know what the prize costs me and in how many weeks the card fills.
Day 0
Ten minutes. Name, stamps, reward, your colours. Add a head-start stamp: a card that is born started gets finished more often than a blank one, and it costs you no margin until the end.
The usual mistake: Leaving it in draft “until it is perfect”. The QR code does not print itself, and the first week brings the most signups.
Done when: The card is published and I opened the link from my own phone.
Day 1
Where the customer pays, at hand height. Print it on something that will not fall over and does not end up behind the till. If you deliver, put the same QR on the box or the bag.
The usual mistake: Sticking it on the entrance door. Nobody stops to scan on the way in; people scan while waiting for change.
Done when: The QR is where payment happens and I have scanned it myself from the customer’s side.
Day 1
The only thing to train is a six-word line said while taking payment: “Have you got our card? Ten seconds.” The rest of the training is scanning once, in front of you.
The usual mistake: Explaining the tool in a meeting. Nobody remembers a meeting; everybody remembers a line they said twenty times.
Done when: Every person on the team has stamped at least one real card.
Day 7 – 14
The novelty wears off and the team stops mentioning it. Look at daily signups: if they fall to zero, it is not that customers are uninterested, it is that nobody is offering it. Say the line in front of the team again.
The usual mistake: Reading the drop as customers not caring. It is nearly always that it stopped being offered.
Done when: I compared week one signups to week two and I know why they changed.
Day 14 – 30
The first customer to fill a card will doubt whether you really mean it. Have your team hand it over fast, without questions, and visibly. That moment is what people tell their friends about.
The usual mistake: Adding small print at the moment of handover. A “only on Tuesdays, actually” turns the best moment of the programme into the worst.
Done when: The team knows exactly what to do when someone shows a completed card.
Day 30
You now have enough data to look at three numbers: how many joined, how many came back at least once, and how many days pass between visits. That third number is the one that says whether the programme is changing anything.
The usual mistake: Looking only at the customer total. Eight hundred emails from people who do not return is not a community, it is a list.
Done when: I know my repeat rate and my days between visits, and I compared them to the benchmarks.
Day 30+
Not at everyone. Pick the drifting group — the ones who have been away longer than normal — and send them something small. It is the campaign with the best return because it goes to people who already knew you.
The usual mistake: Twenty percent for the whole list. It mostly goes to people who were coming anyway, and the ones who matter never hear about it.
Done when: My first campaign went to a segment, not to everyone, and I know how many people it reached.
The templates solve step one, and the benchmarks tell you whether your day-30 numbers are any good.
