Klaviyo Winback Flow: The 2026 Dormant-Customer Playbook

The quiet revenue leak nobody talks about

Roughly 60 to 70 percent of the customers a typical Shopify brand acquires in any given year will not buy again within the next twelve months. That is not a hot take, it is what the Klaviyo benchmark dashboards show for brands doing between 500k and 10M in annual revenue. Most founders obsess over new acquisition costs climbing on Meta while a much larger pool of already-paid-for buyers silently drifts into dormancy.

A winback flow is the single highest-leverage automation you can build after your welcome series and post-purchase flow are live. It costs nothing to send to a list you already own, and every reactivation you earn is a customer you do not have to re-buy through paid ads. This post is the exact playbook we install for D2C clients inside our email marketing service, condensed into a version you can build yourself this week.

TL;DR

-> Segment dormant buyers by days-since-last-order, not by a single "inactive" tag -> Send a 3-email sequence: reminder, incentive, last call -> Ladder the offer based on customer AOV, not a flat 10 percent for everyone -> Sunset non-openers after the flow ends so your sender reputation stays clean

When to run a winback flow

The trigger date is the most debated variable in retention. Run the flow too early and you are discounting customers who would have reordered organically. Run it too late and they have already bought from a competitor or forgotten the brand.

The right trigger window is a function of your product's natural repurchase cycle. Look at your Shopify order data and calculate the median days between first and second order for the last 12 months of repeat buyers. For a consumables brand (coffee, skincare, supplements) that number is usually 35 to 55 days. For apparel it is 90 to 150. For home goods and accessories it can be 180 to 365.

Your winback trigger should fire at roughly 1.8x to 2x the median repurchase interval. A coffee brand with a 45-day median reorder should trigger winback at day 90. An apparel brand at 120-day median should trigger at day 210. This gives organic repeat buyers room to come back on their own before you spend an incentive on them.

Two hard exclusions at the flow level: anyone who has placed an order in the last 30 days (safety net against date-math errors) and anyone currently inside an active checkout abandonment or browse abandonment flow. You do not want a winback email landing while a customer is mid-purchase.

Segmenting by last-order days: the 30-60-90 winback ladder

Most winback flows we audit send the same message to every dormant customer regardless of how long they have been gone. That is a mistake. A customer who last bought 95 days ago has a completely different reactivation probability than one who last bought 400 days ago, and the copy, offer, and subject line should reflect that.

We call our segmentation structure the 30-60-90 winback ladder, named for the three recency tiers you should build as separate Klaviyo segments off your trigger date:

Tier 1 (Recent dormant): 1x to 1.5x median reorder interval past the trigger. These are customers who slightly missed their window. They remember you, they probably still have product at home, and they respond best to a soft nudge. Reactivation rates here routinely hit 8 to 14 percent with zero or minimal discount.

Tier 2 (Mid dormant): 1.5x to 3x median reorder interval past trigger. The middle tier. They need a reason to come back. Expect reactivation rates of 3 to 6 percent with a modest incentive.

Tier 3 (Deep dormant): 3x+ median reorder interval, up to 18 months since last order. Last chance to bring them back before sunset. Reactivation rates here are usually 1 to 3 percent. Anything above that is a win, but the goal is equally about list hygiene.

Customers past 18 months with zero engagement should not enter the winback flow at all. They go straight into sunset, which we cover below.

The 3-email structure

Three emails, spaced 4 to 6 days apart. Not a 7-email epic, not a single "we miss you" blast. Three. Each one has a specific job and you should not blur them.

Email 1: The reminder (Day 0)

No incentive. This is the "hey, remember us" email. Subject line should reference a specific product they bought or a recent brand moment, not a generic "we miss you." Examples that consistently outperform:

• "The [product name] restock just dropped"
• "It has been a minute, [first name]"
• "Your [product] is probably running low"

Body content is short. Three paragraphs max. Remind them what they bought last, show the product they are most likely to reorder based on purchase history, and include social proof (a recent review, a press mention, a new customer count). No discount code. The job of email 1 is to see who comes back without one, and those customers are gold because they did not cost you margin.

If you have not already built a post-purchase flow to educate new buyers, that is a prerequisite for winback working at all. Our guide on post-purchase emails that drive reorders covers the setup.

Email 2: The incentive (Day 4 to 6)

Now you introduce the offer. The offer size depends on the tier (see ladder below) and the customer's historical AOV. The subject line should be specific about what is inside, not coy:

• "15 percent off, just for you, expires Friday"
• "A gift to get you back"
• "Here is that discount we mentioned"

Body should lead with the offer, then back into the product recommendation. Include a visible countdown or an expiration date in the copy. Urgency on winback is real: without it, the email gets saved for later and never acted on.

One Klaviyo-specific tactic: use dynamic coupon codes, not a single static code. A static code shared on a deals site can erode margin fast when dormant buyers who were going to reactivate anyway stack it with a public code.

Email 3: The last call (Day 9 to 12)

Same offer as email 2, but framed as final. Subject lines that work:

• "Last day for your 15 percent off"
• "This expires tonight"
• "We will not send another one"

The "we will not send another one" framing matters because it is true: after this email, customers who did not engage move to sunset. Being honest about that increases the open rate on this email by 20 to 30 percent in most accounts we have tested.

Incentive ladder by AOV

Flat discounts are lazy and they cost you margin. A customer whose historical AOV is 180 does not need the same 20 percent off that a 35 AOV customer needs. Ladder the offer by value tier and recency tier together.

Two principles behind this table. First, high-AOV customers value exclusivity and gifts more than discount percentage, because the absolute dollar discount is already large. Second, low-AOV customers are more price sensitive and respond better to percentage-off framing than to gifts.

If you are not sure how to calculate customer-level AOV inside Klaviyo, the Historic Customer Lifetime Value and Average Order Value profile properties are already computed for you when the Shopify integration is set up correctly. Use conditional splits inside the flow to route customers to the right email variant based on those properties.

For a deeper look at how these decisions ladder into overall LTV, our breakdown of ecommerce customer lifetime value is the companion piece.

Sunset rules: protecting sender reputation

This is the part most founders skip and it is the part that matters most over a 12-month horizon. Sending to dormant unengaged profiles tanks your deliverability. Gmail and Apple both treat "email sent to inactive recipient" as a soft negative signal, and over time your sender score degrades, meaning even your engaged subscribers stop seeing your campaigns in the primary tab.

Build a sunset flow that runs in parallel. The rules:

Automatic sunset triggers:

• Zero opens in the last 90 days AND received at least 8 campaigns
• Zero clicks in the last 180 days
• Completed the winback flow without opening or clicking any of the 3 emails
• Hard bounced once, or soft bounced 3+ times

When a profile hits any of these conditions, move them to a "Graveyard" list and suppress them from all future sends except one final sunset email asking them to click to stay subscribed. If they do not click within 10 days, they are suppressed permanently.

The math on this is non-intuitive. Removing 15 percent of your list usually increases total revenue from email within 60 to 90 days because inbox placement improves across the remaining 85 percent. We have seen clients add 8 to 12 percent revenue lift from a sunset-only change.

This is also why sending winback to 18-month-plus dormant profiles is counterproductive. They have almost no reactivation probability and they are actively hurting delivery of your campaigns to the customers who would buy. If you only do one thing from this post, build the sunset flow.

Measuring reactivation

Three numbers tell you if the winback flow is working. Track them in Klaviyo Flow Reports and in a Shopify cohort view side by side.

1. Reactivation rate by tier. Of the profiles that entered the flow this month, what percentage placed at least one order within 30 days? Benchmark: 8-14 percent for Tier 1, 3-6 percent for Tier 2, 1-3 percent for Tier 3.

2. Revenue per recipient (RPR). Total flow revenue divided by total profiles that entered. This is the honest number. Anything under 0.50 means the flow is underperforming and the issue is usually either offer size or trigger timing. Good winbacks run 1.50 to 4.00 RPR across the full flow.

3. Second-order rate post-reactivation. A reactivated customer who buys once and disappears again is not really reactivated. Track whether customers who come back through winback place a third order within 90 days. If that number is below 25 percent, the problem is your post-purchase experience, not your winback. See our breakdown on building a post-purchase experience that creates repeat buyers for fixes.

If you want to benchmark your overall flow stack, the flows that actually move revenue lists the 7 we consider mandatory and where winback ranks in priority.

Weekly actions to run this playbook

-> Monday: Pull median days-between-orders from Shopify and recalibrate your winback trigger if it has drifted more than 15 percent from the current value -> Tuesday: Review last week's winback flow report. Flag any tier where RPR dropped below benchmark for two consecutive weeks -> Wednesday: A/B test one subject line per email in the flow. Only one variable per test, minimum 500 recipients per variant before calling a winner -> Thursday: Check sunset flow numbers. How many profiles suppressed, how many resubscribed via final click -> Friday: Review reactivated customers who have now placed a second order. What did they buy? Feed that data back into email 1 product recommendations

FAQ

How long should the whole winback flow run?

12 days from trigger to last email. Any longer and the offer urgency decays. Any shorter and you are crowding the inbox.

Should I exclude customers who are in a subscription?

Yes, absolutely. An active subscriber is not a dormant customer, even if they have not placed a one-time order in 90 days. Your flow filter should exclude anyone with an active subscription in Recharge, Shopify Subscriptions, or whatever tool you use.

What if my product has a 12-month repurchase cycle (mattresses, high-ticket goods)?

Winback still works but the mechanics shift. Trigger at roughly 1.5x the median instead of 2x, lean harder on referral incentives (refer a friend and get a discount on an accessory) rather than direct repurchase discounts, and extend the gap between emails to 7 to 10 days.

Can I use SMS alongside the email winback?

Yes, but only for Tier 2 and Tier 3, and only for profiles with explicit SMS consent. SMS on winback converts 2 to 3x email but it also feels more intrusive to a dormant contact. Send one SMS timed between email 2 and email 3.

What do I do with customers who reactivate?

Move them out of the dormant segment immediately, exclude them from winback for at least 120 days, and route them into your standard post-purchase flow. The worst thing you can do is trigger winback again 60 days after reactivation because your date logic is off.