Topic
Almost every business has hundreds of people who bought once and never came back. Usually because nobody gave them a reason. Winning them back costs less than finding new customers, as long as you choose well who to write to and measure how many come back because of the message.
It depends on how often a typical customer comes back. We calculate the median interval between purchases and mark as inactive anyone who has gone twice that long without returning.
Audience Median interval Inactive after
Diners 38 days 76 days
Returning guests 11 months 22 monthsRecommendation. A fixed range ("3 to 12 months") confuses a dormant diner with a guest who was not due back yet.
Three emails. The first one reminds, the second gives a reason with a date and the third asks whether the person wants to keep receiving messages.
Email 1 · day 0 "Three months ago you had dinner with us. The menu has changed."
Email 2 · day 4 Seasonal menu, Tuesdays and Wednesdays only, until the 30th
Email 3 · day 9 "Should we keep writing to you?"
No click or reply → removed from regular sendsRecommendation. Removing people who do not respond protects deliverability for the whole base. Rip Curl did it this way: if a lapsed customer did not react, they only received the important announcements.
A random 10% of the segment is held out and receives nothing. The difference between the two groups is what the sequence generated.
Got the sequence 8,640 came back in 30 days 268 3.1%
Control group 960 came back in 30 days 12 1.3%
Effect 1.8 pts × 8,640 ≈ 156 customers who came back because of the messageRecommendation. Without the control group, 268 would have been counted. The honest figure is 156, and that is what you use to decide whether the offer pays off.
| Segment | What to offer | Why |
|---|---|---|
| At risk: used to spend a lot and are drifting away | An exclusive benefit, no discount | Protects margin with people who pay full price |
| Bought only once | A reason for a second visit: double stamp or a new dish | The second visit is what builds the habit |
| Dormant: twice their interval | An offer with a deadline | They need a concrete reason and a date |
| No clicks in six months | A stay-or-go question, then unsubscribe | Protects deliverability for everyone else |
Case analysis
The fair-trade menswear brand used email to talk again to people who had already shown interest.
Why it worked. It wrote to each person when it had something specific for them, and it tested before giving away margin.
Analysis of a public case. North Marketing did not take part in this work. The figures are the ones published by the vendor (Klaviyo).
Sources: Klaviyo
30 minutes to see where your business is losing customers and what to do first. If you don’t see results in 30 days, we give your money back.
Book your diagnosisWe’ll message you on Instagram with times.