Every store has a quiet group of customers who bought once or twice and then disappeared. They did not complain and they did not unsubscribe. They simply stopped. Winning some of them back is usually cheaper than finding new customers, because they already know the brand, have bought and, at one point, gave you permission to contact them.
But a win-back email is also one of the easiest to get wrong. The “we miss you, here is 20 percent off” message is overused, trains customers to wait for discounts and goes to people whose inbox has not seen you in a year. This guide sets out a more careful approach: how to define lapsed, how to segment, what to send, when to stop and how to measure. It builds on customer lifetime value, RFM segmentation and flow best practices.
Define “lapsed” from your own data
A customer who buys coffee beans every month is lapsed after two months. A customer who buys a mattress is not lapsed after two years. Use your own repeat-purchase intervals. Calculate the median time between first and second orders and between later orders, and define lapsed as a multiple of that, such as one and a half to two times the usual gap.
| Store type | Typical purchase cycle | Rough lapsed point |
|---|---|---|
| Consumables (coffee, supplements, pet food) | 4 to 8 weeks | 2 to 4 months |
| Fashion and accessories | 3 to 6 months | 6 to 9 months |
| Home goods | 6 to 12 months | 12 to 18 months |
| Durable or high-consideration items | Years | Use other triggers, like seasons or product launches |
The figures are examples of the approach, not benchmarks. Your own data should decide. If you cannot tell, plot the distribution of days between orders and pick the point after which the chance of another purchase falls sharply.
Segment before you send
| Segment | How to identify | Approach |
|---|---|---|
| Past best customers | High spend or frequency in RFM, now lapsed | Personal tone, new arrivals, maybe a note from the founder; consider a call for very valuable ones |
| Repeat but modest | Two or three orders | Relevance first: what is new in what they bought |
| One-time buyers | One order, long ago | Reminder of value, reviews, and a gentle offer later in the sequence |
| Discount-only buyers | Bought only on promotion | Be careful: an offer may just train them again |
| Unhappy customers | Returned, complained or low ratings | Do not send a promotion; consider a service follow-up |
Suppress anyone who has an open support issue, who recently returned an item or who already received another campaign. A win-back that arrives in the middle of a complaint damages the relationship it hoped to repair.
A three-email sequence
| Timing | Purpose | Example subject | |
|---|---|---|---|
| 1. Reason to return | At the lapsed point | Relevance: what is new, what they bought, how to reorder | New since your last order |
| 2. Help or proof | 7 to 10 days later | Answer likely reasons for leaving: reviews, a guide, a policy improvement | How others use [product] |
| 3. Incentive or last call | 14 to 21 days after email 2 | A small, time-limited offer, or a preference check | Before we go quiet |
Notice the order: relevance first, offer last. Customers who return because of a reason are usually more valuable than those who return only for a discount. If you offer an incentive, keep it modest and time-limited, give it only to those who did not respond to the earlier emails and test it against a group that gets none.
What to write
- Lead with something useful: a restock, a new version, a seasonal item, a guide to using what they bought.
- Keep it short. One idea, one button.
- Use purchase history. Refer to the category they bought, not their last name and nothing else.
- Admit the gap lightly. “It has been a while” is fine; guilt is not.
- Offer an easy way to update preferences. Some lapsed customers will respond to a lower frequency, and some want to leave.
- Make unsubscribe easy. It protects deliverability and respects the person.
Examples
| Type | Subject | Preview text | Core line |
|---|---|---|---|
| Replenishment | Time to restock your coffee? | The roast you ordered is back | Reorder in one click with your last settings. |
| New arrivals | New since your last visit | Three pieces that go with your table linen | Here is what we added since March. |
| Feedback | Can we ask you something? | One question, ten seconds | What would bring you back? Tap one. |
| Last call | Should we keep in touch? | Tell us what you want to hear about | Stay on the list or tell us how often you want us to write. |
Sunset the unengaged
The last step of a win-back flow is to stop sending to people who do not respond. Continuing to mail contacts who have not opened or clicked in many months lowers your engagement rates and can hurt inbox placement for everyone. After the sequence, move non-responders to a suppression or sunset list and review it periodically. Some stores send a final “confirm you want to stay” message; others simply stop. Follow the consent and unsubscribe rules in your market and your email platform’s guidance. A smaller, more engaged list is generally better than a large, silent one.
Measure with a holdout
Some lapsed customers would have come back anyway. To estimate the real effect, hold back a random 10 to 20 percent of the eligible group from the campaign. Compare orders and revenue per person over the following weeks between sent and held-out groups. The difference is the campaign’s contribution. If the lift is small, change the message, not the volume.
| Metric | How to read it |
|---|---|
| Reactivation rate | Share of the group placing an order within a window, versus the holdout |
| Revenue per recipient | Compare against the holdout, not against zero |
| Margin after any discount | Incentives reduce profit; include them |
| Second order rate after return | Shows whether reactivated customers stick |
| Unsubscribe and complaint rate | A rise means the audience or message is wrong |
| Deliverability | Watch inbox placement and bounces |
Build it
- Define lapsed from purchase-gap data and create a dynamic segment.
- Add exclusions for support issues, recent returns and other campaigns.
- Split into segments for best, modest, one-time and discount-only customers.
- Build the three-email flow with a holdout split.
- Add a sunset rule for non-responders.
- Review results after a full cycle and adjust.
Our email and SMS marketing flows and customer segmentation services build this on Klaviyo and other platforms, and the welcome series guide covers the start of the same relationship.
Subject lines and sending practice
Win-back subject lines work best when they are plain and specific: name the product or the reason for writing, and avoid guilt or fake urgency. Test two or three styles against each other on a small share of the segment before sending to the rest, and look at replies as well as clicks, because a short reply to a win-back email is often the clearest signal of why a customer left. Send during hours when your customers normally open email, avoid sending the whole sequence to the same people in a single week, and keep your sender name and address consistent so that mailbox providers recognise you. If complaints or unsubscribes rise above your usual level for the flow, pause it and review the audience before changing the copy.
A worked example
A store has 6,000 lapsed customers. It holds out 600 and sends the sequence to 5,400. After 60 days, 4.1 percent of the sent group have ordered against 1.9 percent of the holdout, a lift of 2.2 percentage points, or about 119 extra orders. At an average order of $58 and a 40 percent margin after costs, that is about $2,760 of extra margin, less the cost of the 10 percent offer given to second-stage buyers, perhaps $350. The net is around $2,400 for a few hours of setup. Had the store compared the sent group with zero instead of the holdout, it would have credited the campaign with 221 orders, almost double the true effect. The figures are illustrative.
Mistakes to avoid
- One fixed lapsed definition for all products.
- Opening with the discount.
- Emailing unhappy customers with promotions.
- No holdout, so you cannot tell what worked.
- Never sunsetting, which hurts deliverability.
- Sending too many emails in a short period.
Sources and further reading
Figures and rules in this guide come from the primary pages below. Platform limits and prices change, so open the page before you build on a number.
- Klaviyo Help Center: win-back flow, setting up a lapsed-customer flow and segments
- FTC: CAN-SPAM Act compliance guide, unsubscribe and commercial email requirements
- Google: email sender guidelines, requirements that affect deliverability