How to run a win-back email campaign properly: calculate your real lapse point, segment lapsed customers by value, run a short sequence and know when to stop.
Every win-back email campaign guide starts with the email templates. Wrong end. The two decisions that determine whether the flow earns anything are made before a word is written: when does "lapsed" actually begin for your store, and which lapsed customers are worth the effort. Get those wrong and you're discounting people who were about to buy anyway, or chasing people who were never coming back.
This sits under the customer retention strategy piece, as flow number seven of the eight flows.
Find your real lapse point
The default win-back trigger in most platforms is a flat number of days. Your store has its own rhythm. Pull the distribution of gaps between orders from your data and find the point that covers most repeat purchases, around the 80th percentile of order gaps.
Worked example. A coffee subscriber-style store finds 80% of repeat orders happen within 45 days; the win-back fires at day 60. A furniture store finds the same percentile at 14 months; firing a win-back at day 90 there would discount customers who were never lapsed at all. Same flow, wildly different correct timing, and only your own order data can tell you.
Segment by what they were worth
Not all lapsed customers deserve the same chase:
| Segment | Definition | Treatment |
|---|---|---|
| Lapsed VIPs | Multiple orders or high spend, now past lapse point | Full sequence, your best gesture, possibly a personal note |
| Lapsed one-timers | Single purchase, then silence | Light sequence, lead with reasons not discounts |
| Deep lapsed | Past roughly 2x your lapse point, unengaged with email | One final attempt, then the sunset flow |
The deep lapsed group matters for a defensive reason: mailing them indefinitely erodes the sender reputation that the deliverability piece explains. Win-back has an exit, and the exit is suppression.
The sequence itself
Two to three emails over two to three weeks:
- The check-in. No discount. "It's been a while" plus what's new, best sellers, anything genuinely changed since they bought. A meaningful share of recoveries happen here, at full margin.
- The reason. Social proof, a restock, a new range. For one-timers, address the silent objection: maybe the first product wasn't right, so show the breadth.
- The gesture. If margin allows, the incentive lands here, sized by segment value. VIPs get the better offer; that's loyalty maths, not generosity.
Then stop. A fourth and fifth email don't recover customers; they harvest unsubscribes.
Tips
- Exclude anyone with an active cart or recent site visit; they're not lapsed, they're deciding.
- Ask one question in the sequence ("anything we got wrong?") and read the replies. Lapsed customers are your cheapest research panel.
- Measure recovered revenue per recipient and post-recovery repeat rate, not open rate. A recovered customer who buys once on discount and vanishes again may be breakeven at best.
What to read next
- Email deliverability: protecting the sender reputation win-back depends on
- Klaviyo flows: the eight that earn, in build order
- Lifetime value: deciding which lapsed customers are worth the chase
Frequently asked questions
When is a customer considered lapsed?
When their gap since last purchase exceeds the point covering most of your repeat orders, around the 80th percentile of your own order-gap distribution. It's a calculated number, not a default.
Should win-back emails include a discount?
Last, not first. A meaningful share of winnable customers return for a reason rather than a price, and they return at full margin.
How many win-back emails should I send?
Two to three, then suppress non-responders via a sunset flow. Persistence past that point costs deliverability and earns almost nothing.
How Qwrki fits
Qwrki runs retention flows as part of the operating layer: the lapse-point calculation, the segmentation, the build, and the suppression rules that keep deliverability intact. We don't treat win-back as a one-off send; we run it as a maintained part of the lifecycle so the right customers get chased and the rest get a clean exit. Book a call and we'll read your order-gap data to find where your real lapse point sits.
You may also like
- retention
Abandoned cart recovery: the flow everyone has and almost nobody tunes.
Cart abandonment averages just over 70%. Most stores run one default email against it. Here is the full recovery sequence, the timing, and what to tune first.
4 min - retention
The eight Klaviyo flows that cover most of lifecycle revenue.
The eight Klaviyo flows every ecommerce store needs, in build order: welcome, abandonment, post-purchase, win-back and more, with timing and tuning notes.
4 min