Win-Back Campaign: How to Bring Lapsed Customers Back

A win-back campaign re-engages customers who stopped buying. Learn how to define lapsed, segment, sequence messages and measure with a control group.

Win-Back Campaign: How to Bring Lapsed Customers Back

A win-back campaign is a planned series of messages aimed at customers who used to buy from you and have stopped. It works best when you define "lapsed" by each customer's normal buying rhythm, split them by past value, and measure the result against a control group.

This guide walks through how to do that in a way your CRM team can start on this week.

What is a win-back campaign?

A win-back campaign (sometimes called a reactivation campaign) tries to bring back customers whose last purchase was long enough ago that they are unlikely to return without a nudge.

It is different from a general newsletter in three ways:

  • The audience is defined by silence. Only customers who have crossed your lapse threshold are included.
  • The message acknowledges the gap. It speaks to someone who knows you, not a stranger.
  • It has an end. If the customer does not respond, the sequence stops and they are moved out of regular sending.

Win-back is one of the core levers of CRM growth, the work of increasing revenue from customers you already have. If you want the wider picture first, our guide to What is CRM? covers the basics.

How to define a lapsed customer

The most common mistake is using one number for everyone, such as "no purchase in a set number of days". A customer who buys groceries weekly and a customer who buys a winter coat once a year have very different rhythms.

Use this decision rule instead:

  1. Work out the typical gap between purchases for each product category or customer group. Use the median (the middle value), because a few unusual customers can distort the average.
  2. Set "at risk" at a little beyond that typical gap, when the customer is late but still likely to return.
  3. Set "lapsed" at a clearly longer point, when most customers who were going to return have already done so.

For example, a cosmetics brand may find that skincare buyers reorder on a regular cycle, while gift-set buyers appear mainly around holidays. The brand would then run two lapse definitions, not one.

Do this first: pull a purchase-gap report by category and check how many customers sit in the "at risk" and "lapsed" bands. If the lapsed group is tiny, your effort belongs in prevention (at-risk messaging), not win-back.

Segment lapsed customers before you write anything

Not every lapsed customer deserves the same effort. Two questions sort them quickly: how much were they worth, and how likely are they to return?

At a glanceWho gets what kind of win-back
Likelihood to return: low → high
Light nudgeLow value, likely to return. Automated reminder with relevant products, no discount.
Personal reminderHigh value, likely to return. Tailored picks and early access, small or no incentive.
One last tryLow value, unlikely to return. A single message, then suppress.
Invest and askHigh value, unlikely to return. Ask for feedback and test a stronger offer.
Past customer value: low → high

Match effort and incentive to customer value and likelihood of returning.

Likelihood to return can be estimated from simple signals: recent site or app visits, email opens or clicks, items left in a basket, or a past habit of coming back after a gap.

Find out why customers left

A win-back message is only as good as your guess about why the customer stopped. Check these likely causes before choosing the creative:

  • Natural end of need. The customer bought something they only needed once. Win-back will not help much; cross-sell may.
  • A bad experience. Late delivery, a return problem or a support complaint. Cross-check support and delivery data, and fix the issue before you ask for another order.
  • Price or competition. Visible in browsing without buying, or in coupon-only purchase history.
  • Drift. The customer simply forgot you. This group responds best to a relevant reminder.

For example, a fashion retailer might notice that many lapsed customers had a return as their last order. That group should get a message about the returns process and fit guidance, not a generic "we miss you" email.

Build the win-back sequence

A good sequence is short, escalates gradually and ends cleanly. Here is a simple structure you can adapt.

At a glanceA simple win-back sequence
  1. 1Relevant reminderShow products linked to what they bought or browsed. No discount yet.
  2. 2Value or newsShare something new: a launch, a service change, or a useful guide.
  3. 3Incentive, if neededOffer a time-limited benefit, only to segments where margin allows it.
  4. 4Last callA clear, honest final message that the customer will hear less from you.
  5. 5Suppress or move onCustomers who never respond leave regular sending and enter a quiet list.

Start with relevance, escalate only if the customer stays silent, then stop.

Only use channels where the customer has given permission, and follow the data-protection and messaging rules that apply to you. Within that:

  • Email is usually the first step, because it carries richer content at low cost.
  • SMS, WhatsApp, web push and app push suit customers who have engaged with those channels before and need a short, timely prompt.
  • On-site and in-app messages reach customers who return to browse without buying. Personalise what they see using their past purchases.

When channels are run separately, customers can receive the same offer three times in one day. Plan the sequence as one journey with shared rules, a topic we cover in our post on customer journey orchestration.

Decide when to use a discount

Use this rule: relevance first, incentive second, and only where the margin supports it.

  • Give no discount to customers who are likely to return. A reminder is enough.
  • Offer a limited incentive to high-value customers who stay silent after the first messages.
  • Avoid repeating the same offer to the same customers, or they will learn to wait.

Measure win-back with a control group

Some lapsed customers would have come back without any message. If you count every returning customer as a win, you will overstate the result.

The fix is a control group: a random slice of the lapsed segment that receives no win-back messages. You compare its return rate with the group that was messaged. The difference is the effect your campaign actually produced.

Track these measures:

  • Reactivation rate: share of messaged customers who purchased again, compared with the control group.
  • Revenue per customer messaged, including those who did not return.
  • Margin after incentives, so a discount-driven return does not hide a loss.
  • Second purchase: whether won-back customers buy again, which shows if the return was lasting.
  • Unsubscribes and complaints, as a check on over-messaging.
At a glanceBefore you launch
  • Lapse threshold set per category, not one rule for all
  • Lapsed customers segmented by value and likelihood to return
  • Likely reason for leaving checked in support and delivery data
  • Channels limited to those with customer consent
  • Discount rules agreed with finance, with a margin floor
  • Random control group excluded from all messages
  • End-of-sequence rule: suppress non-responders

If you can tick every line, your win-back journey is ready to go live.

Common win-back mistakes

  • Treating the whole base as lapsed. This floods active customers with "we miss you" messages.
  • Leading with a big discount. It cuts margin and may bring back only bargain hunters.
  • Ignoring service problems. A message cannot fix a poor delivery experience.
  • Never stopping. Sending forever to non-responders harms sender reputation and makes results look worse.
  • Skipping the control group. Without it, you cannot defend the budget.

How Boost Up helps

Boost Up Solutions is a CRM and growth agency based in Istanbul. We have worked with 100+ brands, built 1,000+ personalisation campaigns and run seven channels from one place: email, SMS, WhatsApp, web push, app push, in-app and on-site personalisation.

For win-back work, the two parts of our team fit together. On the CRM agency side, we handle strategy, channel operations and customer journey automation. On the CRM growth agency side, we focus on retention, repeat purchase, customer value and control-group experiments.

We are platform-independent: we apply our strategy, segmentation and measurement approach on whichever CRM, CDP or marketing automation platform you already use. Clients have reported results such as a 310% increase in engagement rate at Poca and a 26% higher conversion rate at S Sport Plus.

Conclusion

A win-back campaign works when it is specific: a lapse definition based on each customer's rhythm, segments sorted by value, a short sequence that stops, and a control group that shows what the campaign really achieved.

If you are unsure whether win-back is your biggest opportunity right now, take our 2-minute CRM maturity quiz to see where your CRM stands, or book a free intro call and we will look at your lapsed segments together.

Frequently asked questions

What is the difference between a win-back and a re-engagement campaign?

Re-engagement targets customers who are going quiet but are still active, for example people who stopped opening emails. Win-back targets customers who have stopped buying altogether. The two often run one after the other.

Should a win-back campaign always include a discount?

No. Discounts teach customers to wait for offers and reduce margin. Start with relevance, such as new arrivals in a category they bought from, and keep incentives for high-value customers who do not respond.

Which channel works best for win-back?

It depends on the customer's consent and past behaviour. Email is the usual starting point because it is cheap and rich in content. Test SMS, WhatsApp or push for customers who have engaged with those channels before.

How often should I run win-back campaigns?

Run them as an always-on automated journey that enrols customers as they become lapsed, rather than as a one-off blast. Review the rules and the creative regularly.

What do I do with customers who never come back?

After your sequence ends, move them to a suppression list so you stop sending to them. This protects your sender reputation and keeps your reporting honest.

Want more revenue from your CRM?

Let's look at your current set-up together and find your biggest opportunities in a free 30-minute call.

Take the CRM quiz · 2 min