Your Loyalty Program's First 90 Days: A Launch Plan

Most loyalty programmes are not abandoned because the rewards were wrong. They are abandoned in week three, when the novelty has worn off, the staff have stopped mentioning it, and nobody has looked at a number yet. The first ninety days decide whether you end up with an asset or a forgotten setting in a dashboard.

Here is a plan that assumes what is actually true of a small business: you have very little time, the programme is not your job, and whoever is on shift will decide whether it succeeds.

Before you launch: three decisions, one hour

Do not design a programme. Decide three things and move on — you will learn more in the first month of live use than in a week of planning.

• One mechanic. Stamps if your purchases are frequent and similar in value, points if basket sizes vary a lot, prepaid packages if you sell high-value appointments. Do not combine mechanics at launch; you cannot explain two things at a till.
• One reward, reachable in about a month of normal visits. If a typical customer comes weekly, set the goal at five or six visits, not twelve. A reward nobody reaches teaches customers the programme is not worth attention.
• One sentence your staff will say. Write it down and stick it by the till. "Want a free coffee after five? Scan this, takes ten seconds." If it takes longer than that to say, it will not be said.

Pick a reward that costs you margin rather than capacity — an add-on, a retail item, an upgrade — not the core service that occupies your busiest hour.

Week 1: enrolment is the only metric

In the first week, ignore redemption, ignore revenue, ignore everything except how many customers joined. Nothing else can happen until people are enrolled.

Put the QR code where the customer is already standing still: at the till, on the counter, on the receipt, on the table. Not on the door, not on a poster behind the counter where they have to turn around. Enrolment happens in the twenty seconds while they wait for a card machine, and that is the window you are designing for.

The number to beat in week one is simple: has every member of staff enrolled at least one customer? If one person has done all of them, you have a training problem, not a customer problem, and it will not fix itself.

Weeks 2–3: the dip, and what to do about it

Enrolment always drops in week two. The novelty is gone for your staff, the reminder has faded, and mentioning it now feels like effort. This is the single most common point of failure, and it is a staff problem every time.

Two things work. First, make it visible to staff rather than to customers — a number on the whiteboard in the back, updated weekly, is more effective than any customer-facing poster. Second, remove the friction you did not notice at launch: if adding a stamp takes four taps, or the person on shift has to borrow the manager's login, it will be skipped during exactly the busy hours when your best customers are in.

By the end of week three you should know your realistic enrolment rate per day. That number, not a projection, is what everything else is built on.

Weeks 3–4: earn the bonus, if your platform offers one

If your provider extends the trial when you hit an activation milestone, treat it as the actual goal of month one rather than a nice-to-have. Sharaftona adds 30 free days when you enrol 15 customers in your first 14 days, which is deliberately set at roughly the point where a programme starts producing data worth reading.

Fifteen customers is not a marketing target, it is about two a day for a week. If you cannot reach it, that is genuinely useful information: it usually means the QR code is in the wrong place or the staff sentence is too long, and both are fixable in an afternoon.

Weeks 5–8: the first campaign, and the first real data

By now some customers are partway to a reward and some have gone quiet. That is enough to do something useful.

Send one notification. Not a weekly newsletter — one message, to a specific group, with a specific ask. The two that reliably work first:

• To customers close to their reward: tell them how close. "One more visit and your next coffee is on us" converts better than any discount, because the customer has already invested.
• To customers who have not visited in 30 days: a single reactivation message with a reason to come back this week. Costs nothing to send, and it is the clearest test of whether the channel works at all.

Then look at whether the message moved anything within seven days. If it did, you have a working retention channel and the rest is repetition. If it did not, change one thing — timing, the offer, or the segment — and try once more before concluding anything.

Weeks 9–12: judge it on three numbers

At ninety days you have enough data to make a real decision. Three numbers tell you almost everything:

• Enrolment as a share of transactions. Below roughly one in ten, the programme is not visible enough at the counter to ever matter.
• Repeat rate among enrolled customers versus everyone else. This is the number that justifies the programme's existence, and the only one worth quoting to yourself when you wonder whether it is working.
• Redemption rate. Near zero means the goal is too far away — shorten it. Very high with no change in visit frequency means you are discounting people who were coming anyway, so raise the threshold or change the reward.

Change one variable at a time from here. Businesses that adjust the reward, the threshold and the messaging together never learn which one mattered, and end up concluding that loyalty programmes do not work for them.

What not to do in the first 90 days

Three tempting mistakes, in order of how much damage they cause: launching tiers before you have enough customers for tiers to mean anything; sending something every week until customers mute you; and rebuilding the programme in month two because month one was slower than you hoped. Month one is always slower than you hoped. The programmes that work are the ones still running, unchanged, in month four.

Sharaftona gets a card live in about twenty minutes and includes notifications rather than billing per message, so the ninety days above cost you attention rather than budget — which is the resource that actually decides whether a loyalty programme survives.