Most loyalty programs are launched in an afternoon and then left alone. The first month is where nearly all the fixable problems appear, and almost none of the meaningful numbers do.
Before you announce anything
Settle these first. Every one will be asked at the counter in week one, and staff deciding them individually under pressure is how a program becomes inconsistent.
One stamp per visit or per item? Does a small purchase earn the same as a large one? Can somebody stamp twice in a day? Do stamps expire, and after how long? Does it work at your other branch? What happens if someone loses access?
Write the answers on a card and keep it by the till.
Then decide the staff sentence. Not a briefing, the exact words. Something like: "Are you collecting stamps? Ten gets you a free coffee, takes ten seconds."
Week one: your regulars only
Do not advertise. Spend the first week signing up the people already walking through the door.
Regulars convert at a much higher rate than strangers, they will actually redeem, and their questions surface every operational problem while the volume is small enough that fixing it is invisible.
Watch three things daily:
How long signup actually takes during a busy period. Time it with a stopwatch rather than estimating. If it runs past about thirty seconds, fix that before anything else, because staff will quietly stop offering it at exactly the moments that matter.
Whether staff are saying the sentence. Stand near the till and count.
What customers keep asking. Two or three questions will come up repeatedly, and those become the answers on your card.
Week two: fix the friction, keep going
By now you know where the process breaks. Common findings at this point.
The QR code is in the wrong place. It usually needs to be beside the card reader rather than on a wall, because that is where customers are already looking.
One member of staff has become the person who does loyalty. That is a single point of failure and it means your signup rate is really a rota.
Signup asks for something people hesitate over. A phone number will cost you conversions that an email address will not.
Keep the ask consistent and resist adding anything. The temptation in week two is to launch a promotion. Do not. You are still establishing whether the basic loop works.
Week three: one campaign, one purpose
Now add a single thing, and pick it to match a problem you have rather than because campaigns are available.
If your Tuesdays are dead, double stamps on Tuesday. If new members join and never return, a welcome reward that expires in two weeks. If your regulars are enthusiastic, a referral offer.
One campaign. Short, dated, easy to explain in a sentence. Running three at once means you will not know which one did anything, and it makes the counter conversation complicated at exactly the point you need it simple.
Week four: look at what you have, carefully
Four things are worth looking at, and only two of them mean much yet.
Signups, as a share of transactions rather than a count. This one is meaningful now. Under about a quarter after four weeks of consistent asking means it is not being asked.
Staff friction, gathered by talking to people rather than from a dashboard. Ask what is annoying about it. There will be something.
Redemptions. These will be low and that is expected. A ten stamp card at a twice weekly visit rate takes five weeks, so most of your first cohort has not finished yet.
Repeat visit rate. Do not read anything into this yet. The sample is too small and too biased toward people who were already regulars.
Pick one improvement for month two. One.
What not to do in month one
Do not change the reward. You do not have enough information, and reducing a reward later is remembered for a long time.
Do not run several campaigns at once. You lose the ability to attribute anything.
Do not change the rules repeatedly. Staff stop trusting the rules and start improvising, which is worse than any individual rule being slightly wrong.
Do not skip talking to staff. They know where it breaks and will usually not volunteer it.
Do not judge the program yet. The honest verdict arrives in month two or three.
Month two, when the numbers start working
The first cohort completes their cards somewhere around week six to eight. That is when the useful numbers arrive.
Compare visits per member in the two months before joining and the two after. That single comparison tells you whether the program is doing its job, and everything else is context around it.
Check what percentage of member spend you gave back in rewards. It should be near what you designed for, and if it is much higher something is leaking, usually same day repeat stamping.
Then adjust the threshold rather than the reward, because thresholds are precise and rewards are blunt. How to create a loyalty program covers that arithmetic, and rewards that protect margin covers costing it.
Helpful next reads: