Short answer: start a win-back sequence when a customer has gone noticeably longer than your typical time between orders, often around 1.5 to 2 times that gap. Send about four emails over three to four weeks, and if there’s still no response, stop marketing to them (sunset) to protect your sender reputation.
Many guides say “send a win-back email after 90 days.” That only makes sense if your customers reorder roughly every 45–60 days. A coffee shop and a mattress store need very different timing.
Step 1: define “lapsed” for your store
- Export orders for the last 12–24 months.
- For customers with at least two orders, calculate the days between their first and second order.
- Take the median. This is your typical purchase cycle.
- Set the win-back trigger at about 1.5–2 times that number.
| Typical gap between orders | Win-back trigger (example) | Example products |
|---|---|---|
| 30 days | 45–60 days | Coffee, pet food, supplements |
| 60 days | 90–120 days | Skincare, cleaning products |
| 120 days | 180–240 days | Apparel, accessories |
| 365+ days | Event-based instead | Furniture, electronics |
For products people rarely rebuy, a win-back flow is less useful than a cross-sell or care email.
Step 2: the 4-email sequence
| Timing | Message | |
|---|---|---|
| 1 | Trigger day | “It’s been a while”: what’s new since their last order |
| 2 | +7 days | Helpful content or best sellers related to their purchase |
| 3 | +14 days | Incentive, if margin allows |
| 4 | +21 days | “Should we keep emailing you?” with a clear choice |
Our Customer Retention Checklist includes checks for each step.
Discount or free shipping?
Test both, and pick based on margin. Free shipping is often cheaper for low order values. A percentage discount can work better on high-margin products. Avoid offering win-back discounts so often that customers learn to lapse on purpose.
Sunset and suppression
If someone doesn’t open or click anything during the sequence, stop sending them marketing emails. Mailbox providers watch engagement and complaints. Google’s email sender guidelines say senders should keep spam rates below 0.3%, and mailing people who no longer want your email raises that risk. Keep the contact for transactional messages and suppress them from campaigns.
Because Apple Mail Privacy Protection can make opens look higher than they are, use clicks, site visits and orders as the main signs of engagement.
Segment your win-back
- High-value customers may deserve a personal email from the owner instead of a discount.
- One-time buyers often need a reminder of why they bought.
- Customers with a support issue need the issue solved first.
See our segmentation guide for a simple RFM approach.
A worked example: calculating your lapse point
Say you export orders for the past year and find that customers who bought twice had a median of 62 days between their first and second order. Using the 1.5–2x rule:
- Start the win-back at about 93 days (62 × 1.5) for most customers.
- Treat 124 days (62 × 2) as “lapsed” for reporting.
- Treat around 180 days with no orders and no clicks as a sunset candidate.
How to get the number in a spreadsheet:
- Export orders with customer email and order date.
- Sort by customer, then by date.
- For each customer with two or more orders, calculate days between order 1 and order 2.
- Take the median, not the average, so a few very slow buyers don’t skew it.
If you sell items people buy once every few years, such as furniture, use a much longer window and focus on care tips, accessories and referrals rather than “we miss you”.
Templates for the four emails
Email 1: a reminder with something useful
“It’s been a while since your last order, {first name}. Here’s what’s new since then: {two or three real updates}. [See what’s new]”
Email 2: a reason to come back
“Your favourite, {last product}, is still here, and we’ve {genuine improvement, such as restocked a size or added a colour}. [Shop {product}]”
Email 3: the offer (if you use one)
“We’d like to have you back. Here’s {free shipping / {X}% off} on your next order until {date}: {code}.”
Email 4: the honest check-in
“We don’t want to fill your inbox with emails you don’t read. Do you still want to hear from us? [Yes, keep me subscribed] If we don’t hear back, we’ll stop sending marketing emails for now.”
How to measure a win-back properly
Win-back flows are easy to over-credit, because some lapsed customers would have come back anyway. To measure the real effect:
| Metric | How to calculate | Why it matters |
|---|---|---|
| Reactivation rate | Recipients who order within 30 days of entry ÷ recipients | Main outcome |
| Holdout comparison | Same rate for a random 10% who don’t get the flow | Shows the flow’s real lift |
| Discount cost | Value of codes redeemed | Checks the offer pays for itself |
| Unsubscribe and complaint rate | Per email | Shows if the flow annoys people |
| Re-engaged but no order | Clicked but didn’t buy | Candidates for a later nudge |
A holdout group is the simplest honest test available to a small store. Most email platforms support a random split or let you exclude a random segment.
Edge cases
- Seasonal buyers: someone who only buys before Christmas isn’t lapsed in June. Check order month before including them.
- Subscription customers: if they buy through a subscription, a paused or cancelled subscription is the trigger, not days since order.
- Customers with open complaints or recent refunds: exclude them and handle personally.
- Wholesale or B2B buyers: they usually order on a schedule; contact them directly rather than through a consumer flow.
- Gift buyers: they might only buy once a year by design. A gentle reminder before their next likely occasion works better.
After the flow: sunset properly
People who don’t respond to the final email should stop getting regular campaigns. You can keep them for occasional big announcements, but sending to long-inactive addresses hurts inbox placement for everyone. Google’s email sender guidelines stress keeping spam complaints low, and unengaged recipients are more likely to complain.
Subject line ideas by email
| Ideas | |
|---|---|
| 1 | “New since your last visit”, “A few things we’ve added”, “It’s been a while, {first name}” |
| 2 | “{Product} is back in stock”, “Your favourite, improved”, “Time for a refill?” (consumables only) |
| 3 | “Free shipping on your next order”, “A thank-you for coming back”, “{Offer} until {day}” |
| 4 | “Should we keep emailing you?”, “Still want to hear from us?”, “We’ll stop here unless you say otherwise” |
Win-back checklist
- Lapse point calculated from your own order data, not a generic 90 days.
- Seasonal, subscription and wholesale customers excluded or handled separately.
- Exit on purchase and on unsubscribe.
- Holdout group set up so you can measure real lift.
- Final email has a clear “keep me subscribed” link that updates their status.
- Non-responders moved to a sunset segment automatically.
- Any discount code is single-use with an expiry.
- Reviewed after 90 days of running, using orders rather than opens.
Related: see our 2027 ecommerce email marketing calendar.
Related guides
Frequently asked questions
Should win-back emails always include a discount?
No. Start with useful news or a product reason, and keep any offer for the third email. Test whether the offer adds enough orders to cover its cost.
What is a holdout group in a win-back campaign?
A random slice of eligible customers who don’t receive the flow. Comparing their order rate with recipients shows how many orders the emails actually added.
When should I send a win-back email?
When a customer has gone about 1.5 to 2 times your typical time between orders without buying again.
How many emails should a win-back campaign include?
Three or four over three to four weeks is common.
What should I do with customers who don’t respond?
Stop sending them marketing emails (sunset them) to protect deliverability, while keeping transactional emails working.