Winback Email Flows: When a Customer Is Actually Lapsed, and What to Send

A winback email flow asks a lapsed customer to buy again. A customer is lapsed once they've gone clearly past your own normal reorder gap, worked out from your order data, not a platform default. Send one honest email: what they bought with a one-click reorder, plus your strongest proof or an offer you can afford, then stop. Hayes Media, an eCommerce growth agency for DTC and Shopify brands, sets that window from each brand's time-between-orders data, then builds the winback, replenishment, and sunset flows around it.

This piece covers how to find that window, who should get the flow, what the email should say, when a discount is worth it, and how to measure the result in gross profit. If you want the full set of core automations first, start with our guide to the Klaviyo flows every store needs.

What's the difference between replenishment, winback, and sunset flows?

They're three flows with three different triggers. Replenishment reaches a customer who is on schedule, winback reaches a customer who has stopped, and sunset stops mailing people who no longer engage. Mixing them up is how a winback flow ends up losing money.

Replenishment email: Sent just before the product a customer bought is likely to run out. The customer isn't lapsed. They're due. The job is to make the reorder easy before they go looking somewhere else.

Winback email: Sent after the customer's normal reorder gap has clearly passed. The job is one honest attempt to bring back a buyer who has stopped buying.

Sunset flow: Sent to subscribers who no longer open, click, or buy. The job is to stop mailing them.

Here's the mix-up to avoid. A store sends a discounted "we miss you" email at a fixed day count to customers who buy on a longer cycle than that. Those customers weren't lapsed. They were about to reorder, and the store paid for an order it was going to get at full margin.

How do you find your lapse window from your own data?

Measure the gap between consecutive orders for customers who have bought more than once, find the point by which most of those repeat orders have already happened, and start winback after it. Replenishment goes just before the typical gap. Winback goes after the late end of it.

Step 1. Export your orders. You need customer ID and order date for every order, from Shopify or your email platform.

Step 2. Calculate the gaps. For every customer with two or more orders, calculate the days between first and second order. If you have enough data, do the same for later pairs, kept separate. These are the same order gaps our guide to repeat purchase rate uses to time replenishment reminders.

Step 3. Read two numbers. Sort the gaps and read the median, then read a late percentile, such as the day by which 80% of second orders have happened. The 80% is a choice you make, not a benchmark. Pick a cutoff and stick to it so you can compare over time.

Step 4. Split by product. A 30-serving tub and a 90-serving tub won't share a reorder gap. If your range has different usage cycles, run the numbers per product or per category.

Step 5. Set the triggers. Replenishment fires a little before the median gap for that product. Winback fires after the late percentile. Between the two, the customer is due but not lost, so hold the winback.

Step 6. Re-run it when things change. New pack sizes, price changes, and a subscribe-and-save launch all move the gap.

Worked example (made-up numbers). A coffee brand looks at 1,000 customers who have ordered twice. The median gap between first and second order is 34 days, and 80% of second orders have happened by day 52. So replenishment fires around day 28, and winback starts around day 60. If the brand had used a flat 90 days instead, nothing would reach those customers for a month after the point where most of its repeat buyers have already reordered.

The reverse happens too. A skincare device brand with a long cycle that uses a short flat delay mails customers who are on schedule and offers them money off an order they were already going to place. The repeat purchase curve in an eCommerce cohort analysis shows the same pattern from a different angle: how fast each cohort comes back, and how many of them do.

Who should get a winback flow, and who shouldn't?

Winback is for customers who have bought at least once, are past your lapse window, haven't ordered since, and still receive your emails. Everyone else either needs a different flow or nothing at all.

Include: Customers past the window for the product they bought. For customers with several orders, use their own average gap if you have it. For one-order customers, use the store or product gap from the steps above.

Exclude active subscribe-and-save customers. They're on auto-ship. They aren't lapsed, and a winback email can prompt them to rethink a subscription that was working.

Exclude open support issues. A customer with an unresolved refund or complaint needs a person to reply, not a reminder to reorder.

Exclude people who already ordered. Every winback flow needs an exit the moment the customer buys, including on another channel you can see.

Exclude the unengaged. Someone who hasn't opened or clicked anything in a long time belongs in sunset, not winback.

What should a winback email say?

Send one email. Exit anyone who buys before it sends.

Bring them back to what they bought. Show the product they ordered, with a one-click path back to it. A customer who simply forgot needs nothing more than that path.

Lead with your strongest proof. New reviews from customers like them, a reformulation, a fixed packaging issue, a new flavor. Use only changes that really happened. If the numbers in the next section support an offer, this is where it goes, and only at a level your margin can afford.

Ask honestly. Say plainly that it's been a while and ask if something went wrong. Let them reply to a real inbox, so a person reads the answer.

Then stop. Don't add a second or third nudge. Our guide to Klaviyo flows calls winback one honest attempt, and warns that endless winback drips damage deliverability. Anyone who doesn't buy stays on your regular campaigns. They move to your sunset flow only if they also stop opening and clicking.

Write the way the founder would write to a customer they recognize. No guilt and no fake urgency.

Should a winback email include a discount?

Not by default. An offer earns its place only when the extra customers it brings back produce more gross profit than the margin you give away to customers who would have reordered at full price. You can only know that by testing it against a full-margin version, with a holdout that gets nothing, and you should judge it in gross profit, not orders.

Worked example (made-up numbers). A brand has 2,000 lapsed customers, an $80 average order value, and a 60% gross margin, so a full-price order is worth $48 in gross profit. The full-margin version of the flow brings back 3% of them: 60 orders, $2,880 in gross profit. A 20% off version brings back 4.5%: 90 orders. But each order now carries $16 less margin, so each one is worth $32. That's also $2,880.

In that example the discount bought 30 extra customers for no extra gross profit on the first order. It's worth running only if those extra customers go on to reorder at full margin. If they only come back when there's a code, it isn't.

One more rule keeps discounts honest: don't run the same offer on a fixed schedule, because a customer who sees it every cycle has a reason to wait for it.

Should you use SMS in a winback flow?

Only when there's a real event. For customers who have given SMS consent, text nothing in winback until something happens they'd want to know: the thing they bought is back in stock, or their replenishment window arrives. The winback email does the work of earning attention back. Providers bill by message segment, not by message, so a text has to carry news, not a reminder.

Keep it short, name the product, link straight to the reorder, and make opting out easy. The same exclusions apply: no SMS to active subscribers, open support cases, or people who just bought. For more on how SMS fits next to email, see our guide to SMS marketing for eCommerce.

What does the sunset step do, and why does it matter for deliverability?

The sunset step sends one last email to subscribers who no longer engage, then stops mailing anyone who doesn't respond. Google's sender guidelines say that if messages from your domain are frequently reported as spam, future messages from you are more likely to be marked as spam, and that spam reports can lower your domain's reputation over time.

The same guidelines require all senders to keep the spam rate reported in Postmaster Tools below 0.3%, and recommend keeping it below 0.10%. They also suggest senders consider unsubscribing recipients who don't open or read their messages, which is the job a sunset flow does. You can read them in Google's email sender guidelines.

Define "unengaged" with more than opens. Use clicks, site visits, and orders as well. A customer who stopped buying but still clicks is a winback candidate, not a sunset candidate.

How do loyalty and subscribe-and-save fit with winback?

For products people reorder, winback reaches customers after they've stopped. Subscribe-and-save and loyalty programs work earlier: a subscribe-and-save program removes the reorder decision, and a loyalty program gives a customer a reason to come back before they lapse. Hayes Media builds both for clients alongside email and SMS, and our guide to eCommerce retention marketing covers the wider retention program.

How do you measure a winback flow?

Count reactivated customers and the gross profit from their orders, compared with a holdout group that got nothing. Opens and click rates tell you whether emails were seen. They don't tell you whether the flow made money, and attributed revenue can include orders that would have happened anyway.

Reactivated customers: People who got the winback email and placed an order within the window you set for measuring it. Use the same window for the holdout.

Incremental reactivation: The reactivation rate of the flow group minus the rate of the holdout. This strips out customers who would have come back anyway.

Gross profit, net of discounts: Margin on those orders after any offer, not revenue.

What happens next: Whether reactivated customers order again. A winback that brings people back once, on a code, and never again, adds little to lifetime gross profit.

These numbers feed the scoreboard Hayes Media uses for retention work: lifetime gross profit to CAC. Every customer you win back at full margin raises the lifetime gross profit side without spending more to acquire anyone. Our guide to improving LTV to CAC shows how that ratio moves.

When does this not apply?

One-time purchases. If customers buy a mattress or a sofa once, there's no reorder gap to measure. Put the effort into the post-purchase experience and cross-sells instead. These post-purchase email flow examples are a better starting point.

Too little repeat data. If only a small number of customers have ordered twice, the gap will be noisy. Start from product usage, for example a 30-serving tub used once a day, and replace it with your own data once you have enough.

Gift and seasonal buyers. If customers buy once a year for an occasion, the winback should key off that occasion, not a day count.

Subscription-first brands. If most revenue comes from subscriptions, the work is cancellation and churn flows, which is a different build.

Want your lapse window set from your own data?

Hayes Media builds email, SMS, loyalty programs, and subscribe-and-save programs for eCommerce brands, working in Klaviyo for most accounts, with experience in most other email and SMS platforms. On a discovery call we'll audit your winback, replenishment, and sunset flows against your own time-between-orders data, and show you which customers are being mailed at the wrong time. One-time email projects start from $2,000 USD, and retainers for ongoing full management start from $4,000 USD per month, scoped on the discovery call. Book a discovery call. No onboarding fees. No lock-in contracts. No junior marketers.

Sources

Google Workspace Admin Help, email sender guidelines: https://support.google.com/a/answer/81126

Hayes Media, the Klaviyo flows every store needs: https://hayesmedia.co/blog/klaviyo-email-flows

Hayes Media, SMS marketing for eCommerce: https://hayesmedia.co/blog/sms-marketing-ecommerce