Lapsed Customer Win-Back

This journey is worth building when cancellation reports show that customers whose paid membership ended do not come back on their own within the first 30 days, and you have real changes to tell them about by then. The membership end date, the status of the cancellation window and a renewed payment or valid purchase must be tracked on the same customer record; you also need email consent, separate SMS consent and the terms of a limited-time offer. If people leave over an unresolved problem, resolve it first; an email that claims something is fixed when it is not ends the trust for good.

Incremental win-back rate vs control

Win-back is proven only when more former customers renew a paid relationship in the journey group than in the control group; opening the email or tapping the discount code is not a win-back.

How it is counted: Take all former customers assigned at entry to either group as the denominator and those who form a renewed paid relationship or a valid purchase within the same follow-up window after assignment as the numerator; the difference between the groups is the journey's effect. The denominator stays the same whether or not a message was delivered or clicked.

Also track

• Unprompted return before contactThe share who come back during the 30-day wait without any message; when it is large, much of the win-back happens on its own and fewer customers need the message.
• Contribution of the offer stepCount returns after the email and after the SMS offer separately; if returns come only from the discount, revisit the content of the first message about what changed.
• Retention after returnTrack separately, over the same follow-up window, whether won-back customers keep the relationship going; do not report a brief low-value return as a lasting recovery.
• Unsubscribes and SMS complaintsTrack unsubscribes after the win-back email and SMS complaints; a customer who left has less patience for these messages, and lost consent also closes the door to reaching them later.

• Trigger and entryA paid membership ending opens the journey, which waits 30 days before the first touch. Split former customers at random at entry: journey group or a control group that receives no win-back messages; the 30-day wait runs the same way for both.
• Who must not enterA customer whose cancellation window is still active does not enter; the scope is a closed paid relationship, and anyone who is already a customer again within 30 days gets no message. A dormant user who never paid belongs to Inactive After First Open, and a subscriber who ignores messages to Unengaged Subscriber Sunset.
• Timing and channelsAfter 30 days, if the customer has not returned, an email explains what changed since they left; wait 3 to 5 days. If they still have not returned and SMS is permitted, a limited-time return discount goes out by SMS and the journey waits through the offer window; the return is checked before every touch.
• Exit and handoffWhen the customer becomes a customer again at the 30-day check, after the email or within the offer window, the journey ends as customer again and the pending message is cancelled. Without SMS consent, or when the offer window passes with no purchase, no further win-back messages are sent; there is no handoff.
• Frequency and consentEach former customer gets two touches at most: one email and, where permitted, one SMS. The first touch talks about changes and benefits, not a discount; the discount sits only in the second step and never reaches a customer without SMS consent.

1. 1Should the journey group's win-back rate fail to beat the control group, the messages probably reached customers who were coming back anyway; high opens, clicks or redeemed discount codes do not rescue that. Weigh the gap in renewed paid relationships and how long those returns last, nothing else.
2. 2Whatever gap you find is tied to the 30-day wait, this channel order and the length of the offer window. Leave out anyone who never paid and any active customer who merely ignores your messages, since neither is part of this journey's measurement; if you also record a returning customer in another reactivation flow, do not count the same return twice.
3. 3When few memberships end and returns are rare, a significant gap takes a long time to build up. Give the sample size calculator the return rate you expect in the control group and the smallest lift worth detecting; divide the sample per arm it returns by the former customers entering per day to know in advance how long you will wait.
4. 4If you find a customer who entered while the cancellation window was still open, remove them from the measurement and fix the entry rule; a return email to someone in the middle of cancelling costs consent and corrupts the count.