Winback Campaigns: When Should Customers Be Classified as 'Inactive'?

The most commonly cited rule is that a customer who has not purchased or engaged in 90 days is 'inactive' and should see a winback campaign. Run three emails over a few weeks, offer a discount on the last one, and anyone who doesn't engage is suppressed.

It is a clean rule. It is also frequently wrong.

A customer is at risk when they've gone quiet longer than someone like them usually would, not when they hit some arbitrary 90-day mark.

What a Winback Campaign Actually Is

A winback campaign is a short automated sequence targeted against customers who have lapsed beyond their normal purchase cycle, with the intent to recover future revenue from people who already know your brand and have trusted it enough to buy from you at least once.

Winback is cheaper than acquisition almost by default: you already have the customer's email, their history, and their permission to talk to them. A widely cited (if dated) Harvard Business Review analysis put new acquisition at 5 to 25 times more expensive than retention, and the logic still tracks today.

This customer already likes your brand. Your job is just to clear out whatever's stopping them from buying again.

Why the Inactivity Threshold Must Come from Your Data

There is no universally applicable definition of 'lapsed' that is correct across different product categories, price points, and buying behaviors. The right threshold is specific to your store and your customer base.

• Consumables (supplements, coffee, skincare, cleaning products) run out on a schedule, so winback should start right after that window closes, not at a flat 90 days.
• Fashion and apparel are seasonal, so a spring-and-autumn shopper isn't lapsed in December, just off-cycle, making winback campaigns strategic.
• High-ticket goods like furniture and mattresses are built to last, so a sofa buyer isn't buying another one for 12 to 18 months.
• Luxury buyers quiet at 180 days aren't lapsed, they're just waiting for the right occasion, so skip the discount and send value instead.

To set your own threshold, take the median gap between a customer's first and second purchase and multiply by 1.5 to 2x.

The Open Rate Problem in Winback

Winback campaigns often use email open rate as their primary signal of customer reactivation. Did they open the email? Yes, so they are re-engaged.

This logic has been significantly undermined by Apple Mail Privacy Protection.

With Apple Mail Privacy Protection, opens get recorded automatically, whether or not the person actually reads the email, and Apple's clients now account for over 60% of opens (Litmus, mid-2026). So for winback, don't measure success by open rate, use clicks, site visits, and purchases instead.

What actually indicates re-engagement:

• Clicks on a product link, collection link, or offer code. This requires a human action that MPP does not generate.
• Purchases.
• If you can track a site visit back to an email click, that's a real signal someone engaged.
• SMS opt-in or reply. An active response in a separate channel is a strong reactivation signal.

RFM in Winback: Not All Lapsed Customers Are Equal

Segment before you send. Your old Champions deserve a different email than a one-time discount shopper, and someone quiet but still browsing isn't the same as someone gone dark.

Go big on high-value customers: personal picks, a real offer, maybe SMS. Keep it simple for mid-tier: 2-3 emails, one offer. Barely bother with low-tier discount hunters, one email, then let them go.

• High-value customers: go big with personal product picks, a real offer, and maybe SMS.
• Mid-tier customers: keep it simple with 2-3 emails, one offer.
• Low-tier discount hunters: barely bother, one email, then let them go.

Forget industry benchmarks. Watch your own reactivation rate by tier, that's the number that matters.

Sequence Structure and Message Strategy

1. Email 1 (Day 0): No discount, just a warm reminder with a relevant product recommendation, this catches people who were already coming back on their own.
2. Email 2 (Day 3-7): A value-led nudge, new arrivals or a relevant review, with a modest offer if any.
3. Email 3 (Day 10-14): Your best offer, only for non-responders, with a real deadline so it feels urgent, not permanent.
4. Email 4 (optional, Day 18-21): A final "stay subscribed or we'll stop emailing" message, it can win back a few stragglers and sets up a clean suppression.

Building and tuning this manually in Klaviyo is tedious enough that most teams set it up once and never touch it again. Platforms like CueZoe handle the escalation automatically, adjusting timing and offers as each customer responds, which is exactly the kind of ongoing work most teams don't have time to keep doing by hand.

No engagement after 3-4 emails means it's time to suppress, more emails just hurt deliverability without adding revenue.

Incentive vs. No-Incentive Winback

Don't lead with a discount. Some customers left over a bad product, and a coupon won't fix that. Some just forgot, and a simple reminder brings them back. Some had a bad experience and need that addressed first.

Skip the discount at first, it shows you who returns on their own and keeps customers from lapsing on purpose to get a deal. New launches work the same way, no discount needed, just relevance.

Sunset Policies: When to Stop

Suppression isn't failure, it's the right call once mailing stops working and starts hurting your sender reputation. A common sunset window is 120-180 days of zero engagement, adjusted to how often you send.

For winback, suppress after 3-4 attempts with no click or purchase, then send a final "we'll remove you unless you stay" email to close it out cleanly. A few will opt back in, most won't, and that's fine.

Cutting disengaged contacts doesn't shrink your list, it raises the odds your remaining sends get seen and clicked.

Winback and List Hygiene Are the Same Problem

Winback and list hygiene look like separate tasks, but they're really two stages of one process. Winback is the active attempt, you spot lapsed customers and try to bring them back with specific, escalating messages. List hygiene picks up when that fails, removing unresponsive contacts to protect deliverability and keep your engagement data clean.

Do both well and you end up with a smaller, more engaged list, one that's more valuable for deliverability and for every segmentation and personalization decision that depends on accurate engagement signals. A list full of contacts who haven't interacted in 18 months doesn't just hurt deliverability, it corrupts every behavioral model built on top of it.

The 90-day rule was never the real answer, it's just a starting point. The real work is knowing what normal looks like for your customers, catching a lapse before it becomes permanent, and accepting that some percentage of your list simply won't come back, no matter what you send them.