A regular who used to come in every Thursday hasn’t been in for six weeks. Nobody at the venue noticed — there’s no list of “people who used to come Thursdays,” just a general sense that things feel a little quieter lately. The response, when there is one, is usually a blanket 20%-off email blasted to the entire customer list, which reaches the Thursday regular along with everyone who was in yesterday and didn’t need a reason to come back at all. The offer wasn’t wrong. It just went to the wrong 95% of the list.
Why discounting everyone doesn’t fix it
A site-wide discount solves a targeting problem with a blunt instrument. It reaches your lapsed customers, sure, but it also reaches your most loyal ones, who would have come in anyway — meaning you’re giving away margin on visits you were always going to get, just to reach the smaller group you actually needed to win back. Do it often enough and you train your best customers to wait for the next discount instead of paying full price, which is the opposite of what a win-back campaign is supposed to achieve. Worse, a discount blasted to people who were already coming in regardless doesn’t win anyone back — it just quietly erodes the margin on visits that were never at risk.
What actually needs to happen
Know who’s actually gone quiet
This only works if you can see, per customer, when they last ordered — which means every order needs to build a CRM profile in the first place. Once that exists, “hasn’t ordered in three to four weeks” becomes a filterable list instead of a gut feeling, and that list is almost always a small fraction of your total customers, not the whole database.
Send the win-back to that list only
A targeted win-back message — sent automatically once someone crosses that quiet-for-a-few-weeks threshold — reaches exactly the people who need a reason to come back, at the moment they’re actually deciding whether this place is still part of their routine. Everyone else keeps ordering at full price, because they were never the problem.
Make the offer personal, not generic
The message that works isn’t “20% off everything” — it’s closer to “haven’t seen you in a few weeks, here’s your usual with a small thank-you attached.” Referencing what someone actually orders, pulled from their own order history, reads as the venue noticing them specifically, not running a promotion at them.
Let referrals and gift cards do some of the winning back too
Referral rewards and gift cards work alongside this rather than instead of it — a lapsed regular who gets a small nudge to come back is also a good candidate to bring a friend along, with both sides getting something for it, tracked automatically rather than as an honesty-system verbal offer nobody ever redeems.
Watch the list shrink, not just the campaign send
The real measure of whether this is working isn’t how many win-back messages went out this month — it’s whether the same names keep reappearing on the lapsed list next month, or whether they graduate back into the “recently ordered” group and stay there. A shrinking lapsed list, checked over a few months, tells you more about whether the offer and the timing are right than any single campaign’s open rate ever will.
The best win-back campaign looks invisible
If it’s working, most of your customers never see a win-back message at all — because they never lapsed in the first place. The only people getting the offer are the ones who actually needed a reason to come back, which means you’re spending the discount exactly once, on exactly the customer it was meant for, instead of handing it to everyone as the cost of reaching a few. That’s the whole difference between a promotion and a strategy — one goes to whoever’s on the list, the other goes to whoever actually needs it.