Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Glossary term

Win-Back Campaign

Messages aimed at customers who have stopped buying or engaging, sent once they pass their usual gap between purchases.
StrategyprocessUpdated

Definition

A win-back campaign, also called a win-back flow or reactivation campaign, is a set of messages sent to customers who have lapsed, with the aim of getting them to buy, book, log in or renew again. It belongs to the last stages of the customer lifecycle, after onboarding, repeat purchase and retention have failed to keep someone active. The lifecycle marketing guide places it there and gives its job: acknowledge the gap and give a specific reason to return.

Not every lapsed customer is worth pursuing. Some have left the category, some were never going to buy twice, and some simply forgot or drifted to a competitor. Win-back is aimed at the last group.

In practice: lapse triggers and sequence

The trigger. A win-back campaign starts when a customer passes a lapse threshold. That threshold works best relative to the customer's own buying rhythm, or the category's, rather than as a fixed number of days for everyone, because a customer who buys coffee monthly and one who buys a mattress every few years lapse on very different clocks. The repeat purchase and loyalty guide sets the threshold where the chance of a further order drops sharply in the time-between-orders distribution, which is usually earlier than stores assume. Subscription businesses use cancellation or a long inactive spell instead.

The sequence. A good win-back sequence is short and has an end. It opens with what has changed since the customer last bought, or a reminder of what they bought and what goes with it; makes returning easy; adds any incentive late, for those who did not respond; and then stops. The ecommerce email and SMS flows guide covers how the flow is built, and ends it with a move into a sunset flow that reduces sending and eventually suppresses people who ignored it. The lifecycle email programme playbook puts win-back on the lifecycle map with the other flows, so lapsed customers are not left receiving nothing while new ones receive too much.

Who gets what. Segmentation by past value lets the effort match the customer: a personal message or call for high-value customers, an automated sequence and no more for low-value ones.

Offer, no offer, or no contact

The tempting design is a discount to everyone who lapses. It has two problems. A discount sent to every lapsed customer pays for some returns that would have happened anyway. And a programme that answers lapsing with money teaches customers that leaving pays. The customer retention strategy guide puts it plainly: win-back works best when it addresses the reason customers left and worst when it simply offers money.

There are three options for any lapsed customer, not two: send an offer, send a message without an offer, or send nothing. Each has a cost, and the comparison between them is an empirical question for each business.

Measuring recovery against a holdout

Counting every lapsed customer who bought after a win-back message as recovered overstates the campaign, because some of them were coming back regardless. The honest measure compares lapsed customers who received the campaign with a randomly chosen Holdout Group of lapsed customers who did not, and credits the campaign only with the difference. The win-back offer test extends this to three arms, offer, message only and holdout, read on incremental gross profit per lapsed customer, so the value of the contact and the value of the discount are priced separately.

Illustrative example — not a real result

A spice subscription finds who was coming back anyway

Saffrolow Spice Club sent a 30% discount to every subscriber with no order for 30 days, and its report credited the campaign with about 420 reactivations a month. Most members ordered roughly every five weeks, so the 30-day trigger caught many who were simply between orders. The team held back a random 20% of lapsed members from the next three months of sends. Of the held-back members, 34% ordered within 60 days with no contact at all; of those who received the discount, 39% did. The campaign was recovering about 5 points, not the whole 39, while discounting every order. The club moved the trigger to customers past twice their own usual gap, opened the sequence with new blends, and kept the discount for a final message to those who had not responded.

Win-back works on the customers that Churn Rate counts as lost, or on the way to being lost; in transactional businesses with no cancellation event, lapse is how churn is inferred. Repeat Purchase Rate measures the positive side of the same behaviour, and the time between repeat orders is where the lapse threshold comes from. A Holdout Group is what separates the campaign's real effect from customers who would have returned anyway, and Segmentation decides how much effort each lapsed customer is worth.

Common confusions

  • Win-back is not re-engagement of non-openers. It targets customers who stopped buying or using, not subscribers who stopped opening.
  • Every return after the message is not a recovery. Without a holdout, the campaign takes credit for natural returns.
  • A fixed number of days is not a lapse definition. Buying rhythms differ by customer and category.
  • A bigger discount is not a better win-back. It raises cost on customers who would have come back, and teaches lapsing.
  • Win-back is not endless. A sequence that never stops harms deliverability and ignores a clear decision.

More terms to look up

Where this term is used

Related