Every brand has customers who bought once or twice and then went quiet. Win-back campaigns bring some of them back, and for those who will not return, they give you a clean point to stop sending, which protects deliverability. The brands that get this right do not just add a discount. They figure out who lapsed, why, and what would make a return worth it.
The short version: define lapsed based on your own repurchase cycle, segment by value, use a relevant reason to return, layer channels for high-value customers, and finish with a sunset so you stop emailing people who never respond.
Step 1: Define "lapsed" with your data
A fixed number of days does not work for every brand. A coffee brand with a 30-day cycle and a furniture brand with a two-year cycle define lapsed very differently.
- Calculate the typical time between first and second orders, and between later orders
- Set the trigger a bit longer than the typical cycle (for example, 1.5 to 2 times the median gap)
- Use different windows for different product categories
| Cycle | Example trigger window |
|---|---|
| Fast replenishment | 45 to 90 days |
| Moderate purchase cycle | 90 to 180 days |
| Considered or seasonal purchases | 180 to 365 days |
These windows are starting points. Replace them with your own data.
Step 2: Segment before you send
| Segment | Approach |
|---|---|
| High-value lapsed (top spenders) | Personal, relevant, multichannel |
| One-time buyers | Education, social proof, second-purchase incentive |
| Repeat buyers gone quiet | Appreciation, new products, loyalty benefits |
| Discount-driven buyers | Offer-led messaging |
| Full-price buyers | Value-led, no blanket discount |
| Subscribers who canceled | Address the cancel reason. See Subscription Ecommerce |
| Engaged but not buying | Product-focused, new arrivals |
| Fully unengaged | Short sunset path |
Step 3: Pick the right reason to return
A reason beats a coupon. Options:
- New product or collection relevant to what they bought
- Restock of something they liked
- Improvement: a new formula, fit, or feature
- Loyalty benefit: points expiring, a tier perk, early access
- Reminder: time to reorder, based on usage
- Feedback request: a short survey, with a thank-you
- Incentive: free shipping, a gift, or a limited offer, used where margin allows
Match the reason to the segment. Offer-driven customers may respond to a discount, while full-price buyers may prefer a new product story.
Step 4: Build the sequence
Example sequence for a mid-value lapsed customer
| Day | Channel | Message |
|---|---|---|
| 0 | "It has been a while" with what is new and a relevant product | |
| 4 | Social proof and reviews for products they bought | |
| 7 | SMS (if consented) | Short reminder with a link |
| 10 | Incentive or final reminder | |
| 14 | Sunset check | If no engagement, move to sunset path |
Example sequence for a high-value lapsed customer
| Day | Channel | Message |
|---|---|---|
| 0 | Personal note that recognizes their history | |
| 5 | SMS (if consented) | Short, personal-style reminder |
| 10 | Direct mail | Postcard or letter with a relevant reason and a trackable link |
| 17 | Follow-up referencing the mailer | |
| 24 | Final | Last chance, then sunset or low-frequency list |
See Direct Mail for Ecommerce and SMS vs Email.
Step 5: Write copy that sounds human
- Be direct and warm. Skip guilt trips
- Reference something real: a product they bought, how long it has been, what is new
- Keep one call to action
- Make leaving easy. Offer a preference link or an unsubscribe
- Test subject lines on clicks and conversions, not opens. See AI and the Inbox
Step 6: Sunset cleanly
Win-back is also list hygiene. For subscribers who do not engage:
- Send a final "stay or go" email
- Suppress those who do not respond
- Review the policy quarterly
See Email Deliverability for Ecommerce.
Win-back and peak season
Run win-back before BFCM so reactivated customers become engaged segments for the sale, and sunset before the peak to protect your sending reputation. See BFCM Planning Guide.
Measuring win-back
| Metric | Why it matters |
|---|---|
| Reactivation rate | Share of lapsed customers who purchase again |
| Revenue per recipient | Efficiency |
| Incremental lift (holdout) | What you earned beyond people who would have returned anyway |
| Second-order value and later repeat rate | Quality of recovered customers |
| Unsubscribe and complaint rate | Whether you are pushing too hard |
| List health after sunset | Deliverability impact |
Holdouts matter. Some lapsed customers would have come back anyway, and without a holdout you may credit win-back for them.
Common win-back mistakes
- Using the same trigger window for every product
- Opening with a deep discount to everyone
- Sending to people who have asked to be left alone
- Ignoring subscription cancel reasons
- Not defining a sunset rule
- Measuring only attributed revenue
FAQ
When should I send a win-back email?
After a window based on your typical repurchase cycle, usually 1.5 to 2 times the median time between orders.
Should I offer a discount in a win-back campaign?
Not always. Test value-led messages first, and use incentives where margin supports it.
What is a good win-back rate?
It varies widely. Track your own reactivation rate and incremental lift by segment.
Is win-back different from re-engagement?
They overlap. Win-back targets lapsed customers who purchased, and re-engagement often targets subscribers who have stopped opening or clicking.
Should I use SMS or direct mail for win-back?
For high-value customers, layering channels can help, with consent for SMS and a measured test for direct mail.
How ATTN helps
ATTN builds win-back and sunset programs for ecommerce brands across email, SMS, and direct mail, with clear segmentation and incrementality testing. Talk to our retention team about a win-back plan.