Every business has a quiet list of people who used to come in and stopped. They are the cheapest customers you will ever reacquire, because they already know where you are and what you sell. The hard part is noticing them at all.
When is a customer lapsed?
A customer is lapsed when they have been absent for about three times their own normal gap between visits, which means lapse is defined per customer, not as a fixed number of days for everyone.
| Normal pattern | Normal gap | Treat as lapsed after |
|---|---|---|
| Daily commuter | 1-2 days | ~7 days |
| Twice weekly | 3-4 days | ~14 days |
| Weekly | 7 days | ~21 days |
| Monthly | 30 days | ~90 days |
A single 45-day rule applied to everyone will chase your daily regulars six weeks too late and pester your monthly customers while they are still perfectly normal.
What should a win-back message say?
A win-back message should give one specific reason to come back this week, reference the progress they already have, and be short enough to read on a lock screen without opening anything.
- Name the business immediately: they may not recognise the number.
- Reference their actual state: 'you're 7 stamps in' beats 'we miss you'.
- Give one concrete, time-bounded reason to return.
- Keep it under two lines. This is a notification, not a newsletter.
Joe's Coffee: you're 7 stamps from a free one and your card's been quiet. Next coffee's on us if you're in before Sunday.
How often should I send them?
Send once, and if there is no response, stop. A second message converts a small number of additional customers and costs you a larger number of opt-outs, which removes them from every future campaign permanently.
The discipline to send once is what keeps the channel usable. Wallet push works because it is rare; businesses that treat it as a broadcast channel destroy the open rate that made it valuable.
What reactivation rate should I expect?
Expect 10 to 20% of contacted lapsed customers to return within a fortnight for a well-targeted message with a real offer. Anything above that usually means your lapse threshold is too tight and you are messaging people who were coming back anyway.
Which customers should I not chase?
- Anyone who has opted out. This is a legal line, not a judgement call.
- Customers with one visit ever: they were passing through, and there is nothing to reactivate.
- Customers already returning. Check the date before the campaign, not after.
- Anyone lapsed for over a year. Treat them as a new acquisition instead.
When should I send it?
Send a win-back message shortly before the customer would naturally be deciding where to go: mid-morning for a café, late afternoon for a restaurant, and never during your own busiest hour, when the customer cannot act on it and you could not serve them if they did.
Match the day to the visit pattern you are trying to restore. A weekday-morning commuter should hear from you on a weekday morning, not on a Saturday when their routine is different and your message is competing with the rest of their weekend.
One timing rule overrides the rest: never send while the shop is shut. A notification that arrives when the customer cannot come in converts the impulse into a mental note, and mental notes do not buy coffee.
What offer should I use?
The strongest win-back offer is usually the progress the customer already has, not a discount. Reminding someone they are seven stamps into a ten-stamp card outperforms a generic ten per cent off, because it costs you nothing and it reactivates a goal they had already adopted.
When progress alone is not enough, add something small, specific and time-bounded. "Your next coffee is on us if you're in before Sunday" gives a reason to act this week; "10% off" gives a reason to act eventually, which means never.
| Offer | Typical strength | Cost to you |
|---|---|---|
| Reminder of existing progress | Strong | Nothing |
| Bonus stamps added to their card | Strong | Very low |
| One free item, expiring this week | Strong | One item's marginal cost |
| Percentage discount | Weak | Margin on every item bought |
| "We miss you", no offer | Weak | Nothing, and it earns nothing |
Note the pattern: the cheapest offers are the strongest. Percentage discounts perform worst and cost most, because they neither restore a goal nor create urgency: they simply reduce your margin on a visit that might have happened anyway.
Which channel should I use?
Use the wallet push if the customer still has the card installed, because it costs nothing, lands on the lock screen, and needs no marketing consent when it reports a real change to their card. Fall back to SMS only for customers who have removed the card.
The cost difference compounds. A wallet push to 300 lapsed customers is free and repeatable; the same campaign by SMS carries a per-message fee every time you run it. Over a year of monthly win-back campaigns that gap is the difference between a habit and a budget line.
Email is the weakest of the three for this purpose. A lapsed local customer is not reading a promotional email from a café, and the open rates reflect it.
How do I measure the campaign?
Measure a win-back campaign by visits within fourteen days from the customers you contacted, not by opens or taps. A card notification is opened by nearly everyone, which makes the open rate useless as a signal.
- Record the exact list of customers contacted, with the date.
- Wait fourteen days without sending anything else to that group.
- Count how many of them transacted in that window.
- Divide by the list size. That is your reactivation rate.
- Multiply the returners by your average margin per visit to value the campaign.
Ten to twenty per cent is a good result for a well-targeted list with a real offer. Materially above that usually means your lapse threshold is too tight and you are counting customers who were returning regardless, which inflates the number and teaches you the wrong lesson.
What if they opted out?
A customer who has opted out must not be contacted, and there is no marketing judgement to be made about it. This is a legal line under GDPR and ePrivacy rules in the EU and UK, and under comparable regimes in the UAE and the wider Gulf.
Practically, this means the opt-out has to be respected across every list, including any spreadsheet a staff member built. An opt-out honoured in the platform and ignored in a downloaded CSV is still a breach.
The customer can still hold the card and keep collecting stamps. Leaving the scheme and leaving the marketing list are different decisions, and conflating them costs you customers who simply wanted fewer messages.
How often should I run this?
Run the lapsed report monthly and contact each customer once. Because lapse is defined per customer, the list refreshes naturally each month with new names rather than the same people repeatedly.
Keep a simple suppression rule alongside it: nobody receives a win-back message twice within six months. Without that rule, a monthly campaign slowly turns into harassment of the same non-responders, and the opt-outs it generates remove them from every future campaign permanently.