Most ecommerce brands have a winback flow turned on. What most of them have set up is not quite what a winback is supposed to do.
The most common version: a three-email sequence that fires when a customer hasn’t purchased in 180 days, starting with “We miss you!” and ending with a discount. It runs quietly in the background, gets minimal results, and gets credited with some revenue in the attribution dashboard anyway because a small percentage of customers who were going to come back on their own happened to receive an email first.
That is not a winback campaign. That is a late-stage re-permission attempt dressed up as retention.
A properly built winback email campaign is one of the most precise tools in the ecommerce email toolkit. It fires at the right moment in the customer lifecycle – based on your brand’s actual repurchase data, not a generic time threshold – and it creates a structured, escalating reason for a lapsed customer to come back before they’ve fully moved on. The difference between a winback that compounds retention value and one that just fills up a flow report is almost entirely in the strategy layer.
This article covers that strategy layer in full: how to set the trigger correctly, how to build the email sequence, what to put in each message, how to segment across different customer types, and where winback campaigns fit inside a broader retention system.
Key takeaways
- A winback flow trigger should be based on your brand’s average repurchase window, not a generic 90- or 180-day rule. Brands that set triggers at 180 days are reaching customers who have mostly already decided to leave.
- Most winback reactivations happen before the incentive email. Jumping straight to a discount in the first message conditions customers to wait for it – and costs margin on customers who would have returned anyway.
- One-time buyers and repeat buyers need different winback sequences. The same messaging that nudges a loyal customer back can feel generic and off-putting to someone who only bought once.
- Winback is a revenue-driving flow. It is distinct from a sunset flow, which is a list hygiene mechanism and does not belong in the same category.
- Winback campaigns (manual sends to lapsed segments) and winback flows (triggered automations) serve the same strategic purpose but operate differently. Both have a place in a mature retention program.
- The metrics worth watching are returning customer rate and revenue attributed to the winback flow – not open rate or click rate, which don’t tell you whether the campaign is actually working.
What we’ll cover
- What a winback email campaign is actually trying to do
- Why most winback flows fail at the trigger level
- How to calculate the right trigger window for your brand
- The winback sequence – what to include in each email
- How to segment a winback campaign across customer types
- Winback campaigns vs. winback flows – and when you need both
- The incentive question – when to offer one and how to structure it
- What to do with customers who don’t respond
- How winback fits inside a full retention system
- What good looks like – metrics worth tracking
What a winback email campaign is actually trying to do
A winback campaign exists for one reason: to re-engage customers who bought before and have since gone quiet, before that silence becomes permanent.
That framing matters because it shapes everything about how the campaign is built. You are not reaching out to strangers. You are reaching out to people who have already made a purchase decision in your favor – who already know your brand, already trusted it enough to give you their money, and who have, for whatever reason, stopped returning. Your job is not to convince them who you are. Your job is to give them a relevant, well-timed reason to come back.
This is fundamentally different from a cold lead nurture. Lapsed customers are not unaware of your brand – they are disengaged from it. The causes are varied: they found an alternative, they didn’t need to repurchase yet, they had a poor post-purchase experience, they got busier and your brand drifted out of mind. A winback campaign that treats all of these as the same situation will underperform. One that is built with behavioral context – who bought, what they bought, how long they’ve been inactive relative to what’s normal for that product – will perform significantly better.
The economics make this worth taking seriously. Research by Frederick Reichheld of Bain & Company found that increasing customer retention rates by just 5% increases profits by 25% to 95%. Reactivating a lapsed customer who already knows and has bought from your brand is materially cheaper and faster than acquiring a new one from scratch.
The strategic intent is also about timing. Winback campaigns are not a last resort. They are an intervention at the point where a customer’s absence has become statistically notable, but before the disengagement is complete. That window is narrower than most brands assume – and that is precisely why trigger timing is the most important variable in the entire architecture.
Why most winback flows fail at the trigger level
The most common mistake in winback email campaign setup is not the copy, not the design, and not the incentive. It is the trigger.
Most brands set their winback trigger at 90 or 180 days after the last purchase. This is a default inherited from template libraries and agency playbooks that have never been adjusted for the specific brand running them. And for most ecommerce categories, it is far too late.
Here is what actually happens at 180 days. A customer who was going to repurchase on their own has already done so. A customer who was on the fence about coming back has moved on to a competitor, found a substitute, or simply forgotten about the brand. The small segment that remains – the ones who still might respond to a 180-day email – is a thin slice of what was once a reactivatable audience.
The 180-day trigger survives because it generates just enough revenue to look defensible in a dashboard. But what the dashboard does not show is all the customers who lapsed between day 45 and day 120 while the winback flow was waiting to fire. Those customers were reachable. They were not reached.
The underlying issue is that most brands set winback triggers by default, not by analysis. The question to answer first is not “when should we send a winback?” but “at what point does a customer’s absence become unusual given their expected behavior?” That is a data question. It requires looking at your actual order frequency distribution – the average gap between purchases, the median, the tails – and then setting a trigger at the point where a customer who hasn’t come back yet is statistically overdue.
For most ecommerce brands with a replenishable product (supplements, skincare, haircare, pet supplies), that window is somewhere between 45 and 90 days. Brands with longer natural repurchase cycles (apparel, home goods) might appropriately trigger closer to 90 to 120 days. Furniture, large appliances, and categories with genuinely multi-year purchase cycles are legitimate exceptions where a longer delay makes sense.
The 180-day default makes almost no sense for any consumable or regularly repurchased product category. This is exactly the kind of gap our Klaviyo audit checklist for ecommerce stores is built to surface – most brands we audit have their winback trigger set 2–3x later than their actual repurchase data justifies.
How to calculate the right trigger window for your brand
Setting the right winback trigger requires pulling actual purchase interval data from your store. In Klaviyo or Shopify Analytics, this means looking at the distribution of time between first and second orders across your customer base. You want to understand:
- What is the median time between first and second purchase?
- What is the average time between all consecutive purchases for repeat buyers?
- What does the distribution look like? Are most customers clustered around a short window, or is it spread across a wide range?
Once you have that data, the trigger logic becomes clear. If 70% of your repeat customers make their second purchase within 60 days of their first, a customer who reaches day 75 without repurchasing is statistically unusual. That is when the winback flow should fire – not at 180 days.
A practical approach for most brands:
Step 1: Pull all customer purchase intervals from Shopify. Export order data and calculate the number of days between each customer’s consecutive orders.
Step 2: Calculate the median and 75th percentile of those intervals. If the median is 45 days and the 75th percentile is 75 days, that means 75% of your repeat buyers have returned within 75 days.
Step 3: Set your first winback email to trigger around the 75th percentile window. You are catching customers just past the point where most repurchases naturally happen – early enough that the brand is still top of mind, late enough that you’re not interrupting a normal purchase cycle.
Step 4: Adjust by product category if you sell across multiple categories with different repurchase cycles. A customer who bought your highest-frequency SKU (a 30-day supplement supply) needs a different trigger than one who bought a single-purchase item with low repurchase potential.
This analysis also has a downstream effect on customer lifetime value: the earlier you reach a lapsing customer, the lower the cost of reactivation and the higher the probability that the recovered relationship compounds into additional purchases.
This is not complex analysis. It is the kind of work that takes a few hours in a spreadsheet or Klaviyo reporting, and it has a direct impact on how many customers you actually reach while they are still reactivatable.
The winback sequence – what to put in each email
A winback flow for most ecommerce brands runs two to four emails. The number is less important than the logic – what each email is trying to accomplish, how they escalate, and what exits should be in place so customers who convert early don’t keep receiving the sequence.
Email 1 – Re-open the relationship
The first winback email does not lead with a discount. Its job is to re-open the relationship with a relevant, low-pressure reason to return.
What works at this stage: a tone that acknowledges the customer’s history with the brand without being manipulative about it. Something new – a product launch, a reformulation, a seasonal collection – gives a natural reason to reach out that is not “please come back.” Personalized product recommendations based on what the customer previously bought are more likely to land than a generic bestseller grid.
What does not work: “We miss you” subject lines paired with a bland email that links to the homepage. That combination signals automated irrelevance. The customer who receives it knows immediately that no one looked at their purchase history before sending.
The goal of email 1 is not conversion – though conversion at this stage is a welcome outcome. The goal is to re-establish relevance and open the door without burning goodwill or training the customer to expect a discount.
Timing: First email should fire at the trigger point established by your purchase interval analysis.
Email 2 – Build the case for returning
If the customer has not responded to email 1 after three to five days, email 2 digs deeper. This is where social proof, new reviews, product updates, or a “here’s what you’ve been missing” angle works well. For brands with a strong product story (supplements with usage data, skincare with visible results), this is the moment to restate the product’s value in a concrete, results-focused way.
Email 2 is still not the incentive email. The incentive is being preserved for the third position – which means that customers who would have come back anyway will often convert here, before the discount is deployed.
Timing: Three to five days after email 1, if no purchase has occurred.
Email 3 – The incentive
By email 3, the customer has seen two relevant, value-focused touchpoints and has not converted. At this stage, introducing an incentive is appropriate. The incentive acknowledges that there is a specific, limited-time reason to return now rather than later.
The incentive does not need to be a large discount. A relevant offer that matches the customer’s purchase history – for example, a discount on a specific product they’ve bought before, or free shipping on their next order – is more targeted and less margin-intensive than a blanket percentage off.
Timing: Three to five days after email 2.
Email 4 (optional) – The last signal
For customers who have not responded after three emails, a fourth “last chance” message can be appropriate if the brand has high-margin products or a strong enough reason to make one final attempt. This email is explicit: it signals that this is the last email in this sequence, and often uses that transparency as a conversion hook.
After email 4 (or email 3 if you run a shorter sequence), customers who have not responded move into suppression or a re-permission path, depending on their overall engagement history. This transition is not punitive – it is list hygiene. Continuing to send to a confirmed-non-responder damages your deliverability and wastes send volume on an audience that is not engaging.
How to segment a winback campaign across customer types
A winback email campaign that sends the same sequence to every lapsed customer is leaving significant performance on the table. The customers in your lapsed segment are not a homogenous group – they have meaningfully different relationships with your brand, different purchase histories, and different reasons for going quiet.
One-time buyers vs. repeat buyers
This is the most important segmentation split in a winback campaign.
A customer who bought once and never returned is a structurally different re-engagement challenge than a customer who bought four times before going quiet. The one-time buyer may have had an acceptable but not outstanding experience – good enough to buy, not good enough to return on their own. The repeat buyer has demonstrated loyalty and has lapsed despite it. The reasons for their absence are likely different, and the messaging that will resonate is different.
Academic research published in the Journal of Marketing supports this directly: a study by Kumar, Bhagwat, and Zhang (2015) found that the stronger a customer’s first-lifetime relationship with a brand, the more likely they are to accept a win-back offer – and the more profitable they are in their second lifetime. First-time buyers with a thin purchase history require a fundamentally different reactivation approach than high-frequency buyers who’ve gone quiet.
One-time buyer winback: The sequence needs to do more trust-building work. It should remind the customer what made the brand worth buying in the first place, surface reviews and social proof prominently, and – if an incentive is introduced – frame it as a reason to give the brand a second chance rather than just a discount.
Repeat buyer winback: The sequence can be more direct and personal. A customer who has bought five times does not need the brand’s origin story. They need a relevant reason to return – a new product in their category, a value offer tied to their purchase history, or simply an acknowledgment that they’re missed in a way that doesn’t feel automated.
This is also why the post-purchase email flow that preceded the winback matters so much. A customer who received a strong post-purchase experience – good product education, a relevant cross-sell, a well-timed review ask – is more likely to re-engage with a winback touchpoint than one who received a generic order confirmation and nothing more.
High-AOV vs. low-AOV customers
A customer who spent $200 per order warrants a different re-engagement approach than one who spent $35. For high-value customers, the winback sequence can afford to be more generous in the incentive tier – a free gift, a more meaningful discount, or a personal-feeling outreach that does not look like a broadcast. For lower-AOV customers, a clean, direct offer aligned to their previous product is typically sufficient.
Segment by recency within the lapsed window
Not all “lapsed” customers are equally lapsed. A customer who triggered at day 60 (just past the repurchase window) is meaningfully different from one who has been inactive for 110 days. Earlier-stage lapsed customers may need a lighter touch; further-along lapsed customers may need a stronger reason to return.
In Klaviyo, this is straightforward to implement using conditional splits inside the flow, with different paths based on days since last purchase. It is not complex flow architecture – but it is one of the details that separates a well-built winback from a blunt instrument.
Winback campaigns vs. winback flows – and when you need both
There are two ways to execute a winback email campaign strategy, and they are not interchangeable.
A winback flow is a behavioral automation in Klaviyo. It is triggered automatically when a customer reaches the defined lapsed threshold – day 60, day 75, whatever your purchase interval data dictates. It runs 24/7 without anyone hitting send. Every customer who hits that threshold enters the sequence. This is the backbone of any winback strategy at scale.
A winback campaign is a manual broadcast send to a defined segment of lapsed customers. It is typically used for one-time re-engagement pushes – a new product launch that is highly relevant to a lapsed segment, a seasonal moment that creates a natural reason to return, a clearance event, or a deliberate reactivation drive against a dormant portion of the list.
Most brands with a functioning winback program use both. The flow handles the continuous, real-time re-engagement of customers as they individually cross the lapsed threshold. The campaign handles specific moments where there is something new to say to the broader lapsed audience at once.
The strategic difference matters when planning content. A flow message needs to be written for an individual customer in an individual moment – personalization is high, timing is tight. A campaign message is broadcast by nature, which means the reason to reach out needs to be genuinely compelling for the entire lapsed segment, not just a generic “come back” push.
One operational note: if you are running active winback flows, a campaign send to the same lapsed segment needs coordinated timing. A customer who received email 1 of the automated flow two days ago and then receives a campaign targeting lapsed buyers the following day is experiencing two simultaneous winback attempts – which signals that your system does not know what it’s doing. Flow exclusion logic and suppression conditions between campaigns and flows are a prerequisite for this coordination to work correctly.
For a full view of how flows and campaigns sit within the same system, our guide to Klaviyo flows for ecommerce brands covers the complete automation layer – including where winback fits relative to the other flows every mature account should have running.
The incentive question – when to offer one and how to structure it
Incentives in winback campaigns create a genuine tension that most brands resolve by defaulting to a discount. That default is worth questioning.
The case for incentives is real: a meaningful offer gives a lapsed customer a specific, time-limited reason to act now rather than “sometime soon.” That urgency is valuable, especially for customers who are on the fence. Without an incentive, some customers who would have responded to one will not convert.
The case against leading with incentives is equally real. Customers who receive a discount every time they lapse learn to wait for the discount. If your winback flow fires at day 60 and email 1 contains a 15% off code, you have just trained your customer base that loyalty is not rewarded – lapses are. The customers who return consistently without discounts are your most valuable cohort. Front-loading incentives into the winback sequence selectively de-values that cohort.
The right approach is an escalating incentive structure:
- Email 1: No incentive. Relationship and relevance only.
- Email 2: Still no direct incentive. Value restatement, social proof, or a compelling product reason to return.
- Email 3: Introduce the incentive. At this point, the customer has demonstrated that they need an additional push. The incentive is doing its proper job – it is the final mechanism, not the first move.
The size of the incentive in email 3 can vary by customer segment. For high-AOV or high-frequency buyers, a more generous offer is justified by the lifetime value of retaining them. For first-time or low-value buyers, a smaller incentive (free shipping, 10% off) is often sufficient.
One important margin consideration: not all products justify an incentive-driven winback at all. If your margins cannot absorb a re-engagement discount without making the recovered sale unprofitable, a value-based winback approach – social proof, new product education, community elements – is more appropriate than discounting into the red. The metric that should inform this decision is customer lifetime value at the segment level, not the single-order economics of the reactivated purchase.
What to do with customers who don’t respond
After a complete winback sequence – two to four emails over two to three weeks – customers who have not converted or engaged fall into one of two categories:
Category 1: Unengaged but not actively bouncing. These customers have not opened, not clicked, and not purchased. They may still be reachable with a different message or a different timing in the future, but continuing to send regular marketing emails to them starts to erode your deliverability. The right move is suppression from active marketing sends while keeping them in the customer database for future re-permission attempts.
Category 2: Hard disengagement. Customers who unsubscribed or marked as spam during the winback sequence. These contacts should be removed immediately and permanently from marketing sends. Any future contact is counterproductive and potentially legally risky depending on the channel and jurisdiction.
What should not happen: continuing to send these customers into your normal campaign flow as if the winback sequence never occurred. If they are receiving weekly promotional campaigns while simultaneously receiving winback emails, neither sequence is doing its job properly, and your deliverability is absorbing the cost of sending to a confirmed non-engaged audience.
The broader principle here connects to the distinction between winback flows and sunset flows. The winback flow is the revenue-driving sequence that tries to bring customers back. The sunset flow is the list hygiene mechanism that removes chronically disengaged subscribers from your active sending audience. They are sequential, not simultaneous. A subscriber who completes the winback flow without engaging becomes a candidate for the sunset flow – not a recipient of the same winback emails running again six months later.
Understanding the Klaviyo flows for ecommerce that cover each stage of this journey – including where winback hands off to sunset – is what separates a well-architected retention program from one that is running random automations in parallel.
How winback fits inside a full retention system
A winback email campaign does not operate in a vacuum. Understanding where it sits in the broader retention system is what determines whether it is well-designed or just present.
The customer journey context
The winback flow addresses one specific stage in the customer lifecycle: the at-risk lapsing segment. It exists because the post-purchase flow, the cross-sell sequence, and the loyalty program did not create enough pull to keep this customer coming back on their own.
This has a practical implication: if your winback flow is doing a heavy lifting job – if a large percentage of your customers are triggering it – that is not a winback problem. That is a post-purchase problem. The winback flow is catching churn that should have been prevented earlier in the lifecycle. A brand with a strong post-purchase email flow, relevant cross-sell sequences, and a functioning loyalty structure will see fewer customers reach the winback trigger in the first place.
This is the architecture logic behind the full Klaviyo flow stack. Each stage of the customer journey should have automated coverage that makes lapsing less likely before the customer even becomes a winback candidate. Shopify’s own ecommerce research has found that loyal customers generate 44% of total revenue and 46% of orders despite making up just 21% of the customer base – a figure that underlines exactly what is at stake when lifecycle coverage has gaps.
Multi-channel winback
For brands running a broader retention stack beyond email, the winback strategy expands across channels. In a well-coordinated retention system:
- Email carries the primary winback sequence, with the full email architecture (no-offer lead, value build, escalating incentive).
- SMS can be deployed as a secondary touchpoint for customers who have an SMS opt-in. SMS is not the place to run the same sequence in parallel – it should play a distinct role, typically as a shorter, high-urgency message after the email sequence has run without response.
- Direct mail is a high-impact, high-cost reactivation channel that works well for high-LTV lapsed customers – those who spent significantly across multiple orders and have gone quiet. A physical piece of mail stands out in a way that a digital touchpoint cannot, and for premium brands the investment is often justified by the lifetime value of the recovered customer.
- Loyalty programs are a natural winback mechanism for brands that have one. A points expiry reminder, a tier status nudge, or an exclusive member offer is a relevant, non-promotional reason for a lapsed loyalty member to return. The loyalty program provides the reason; the email carries the message.
The key principle for multi-channel winback is coordination. A customer who receives a winback email sequence, a winback SMS, and a direct mail piece all in the same two-week window – without those being deliberately sequenced and designed to escalate – is not experiencing a retention system. They are experiencing channel noise. The channels should be planned as a single coordinated re-engagement arc, with each channel earning its place in the sequence based on the customer’s response (or non-response) at each prior touchpoint.
What good looks like – metrics worth tracking
Measuring winback email campaign performance requires knowing which metrics actually indicate whether the campaign is working – and being clear about the ones that do not.
Metrics that matter
Returning customer rate. The primary outcome metric for any retention effort. If your winback program is working, you should see the overall returning customer rate trend upward over time as more lapsed customers re-enter the active buyer pool. This metric is measured at the account level, not the flow level.
Revenue attributed to the winback flow. Klaviyo’s attribution shows revenue generated by contacts who converted after receiving a winback flow email. This is a meaningful metric – but it should be interpreted with context. Some customers would have returned anyway, and the attribution model will assign them to the flow regardless. The more useful question is whether the total incremental revenue from the winback program justifies the cost of building, running, and optimizing it. For most brands doing serious volume, it does.
Reactivation rate. The percentage of customers who enter the winback sequence and make a purchase before or during it. This is the most direct flow-level measure of effectiveness. Track it by segment (one-time vs. repeat buyers, high vs. low AOV) to understand which audiences the current sequence resonates with and which need adjustment.
Repeat purchase rate by reactivated cohort. Customers who are reactivated by a winback campaign are not all equal. Some will return and become long-term repeat buyers again. Others will take the incentive and disappear. Tracking what happens to reactivated customers over the 90 days after their return purchase tells you whether the winback is recovering genuinely valuable customers or just generating one-off discount transactions. For a deeper look at how to increase repeat purchases in ecommerce, this cohort-level view is exactly where you should be spending time.
Metrics that do not matter here
Open rate. A useful diagnostic signal when troubleshooting subject line performance, but not an accountability metric for the campaign’s business impact. An open rate tells you the email was seen – not that it worked.
Click rate. Same logic. Useful for creative optimization. Not a proxy for revenue impact.
Neither metric tells you whether the winback campaign is actually moving returning customer rate or recovering meaningful customer lifetime value. Those are the numbers that connect to business outcomes.
If an agency partner is reporting winback performance primarily through open rate and click rate, push back. Ask to see reactivation rate by segment and the retention trajectory of reactivated customers over 90 days.
Conclusion
A winback email campaign is one of the most precisely targeted tools in the ecommerce retention toolkit. Unlike a broad campaign or a welcome series, it reaches a specific audience – customers who already said yes once – at a specific moment that matters: the point where their engagement is fading but has not yet gone.
What separates winback campaigns that consistently drive repeat purchase rate improvement from ones that generate just enough attributed revenue to stay on is almost entirely strategic. The trigger needs to be calibrated to your brand’s actual repurchase data. The sequence needs to be built with a deliberate escalation logic that reserves incentives for the customers who genuinely need them. The segments need to account for the real differences between a first-time buyer and a loyal returning customer who has gone quiet.
And the winback flow needs to sit correctly within the full Klaviyo flow architecture – not as an isolated automation, but as one stage in a lifecycle system that gives every customer the right message at the right moment from first subscribe through to re-engagement.
At Retention Side, when we audit an ecommerce brand’s email program, the winback flow is almost always one of the highest-leverage areas we find. Not because winback is the most complex thing to build, but because the gap between what most brands currently have and what a well-calibrated winback can do is substantial – and the fix is usually strategic, not technical.
If you want to understand what a properly built winback sequence would look like for your brand’s specific repurchase data and customer segments, that is exactly the kind of diagnostic we run at the start of every engagement.