Playbook case study

Win-back Case Study: Reactivating Dormant Customers

By Vishwajeet Kumar · 6 min read · Updated 4 November 2026

Quick answerTo win back dormant customers, define dormant clearly, segment by past value, run a short staged sequence across channels with a reason to return, stop when people do not respond, and measure against a control group.

This is an anonymised playbook of the approach, not a client report. The scenario and every number in it are illustrative.

The situation

Scenario: A shopping app with about 80,000 customers who have not purchased in 90 days.

A big group of past customers has gone quiet. A monthly blast to all of them brings back very few and trains the rest to ignore you.

How I diagnose it

Define dormant using your own purchase cycle, not a guess
Split by past value, number of orders and likely reason for leaving
Check consent for each channel and recent message fatigue
Look at what returning customers did first in the past

The plan

Segment by past value

High-value customers get a personal note. Others get a lighter, cheaper touch.

Lead with a reason, not a discount

New arrivals, a changed policy, an easier checkout. Keep offers for the second step.

Run a staged sequence

Three to four touches across channels over about two weeks.

Set a stop rule

After the sequence, pause messages for people who do not respond.

Hold out a control group

Keep 10% unmessaged to measure organic returns.

Learn and refresh

Review which segment and message worked, then refine for the next cohort.

A starting journey you can adapt

When	Channel	Message
Day 0	Email	What is new since they left, with one clear button.
Day 3	Push	A short reminder tied to their favourite category.
Day 7	WhatsApp	A personal note with a limited offer, for opted-in users.
Day 14	Email	Last call, then pause messaging for 30 days.

Numbers to track

Win-back rateRevenue per reactivated customerSecond order rate within 30 daysUnsubscribe and opt-out rateIncremental revenue versus control

Illustrative example

What this looks like in numbers

Illustrative example

A shopping app with about 80,000 customers who have not purchased in 90 days.

The numbers below show the maths of a typical scenario. They are not a client result.

3% to 7%reactivation within 30 days in the scenario

+3,200extra reactivated customers (80,000 x 4 points)

14 dayslength of the staged sequence

Illustrative figures. Your own results will differ. We always measure against a control group and report the real numbers plainly.

Common pitfalls

Training customers to wait for a discount
Messaging the whole dormant base the same way
Counting organic returns as campaign wins
Hammering people who have already ignored you

Frequently asked questions

When is a customer dormant?

It depends on your purchase cycle. A grocery app might use 30 days of silence. A furniture store might use 12 months. Use your own data.

Should I always offer a discount in a win-back?

No. Try a reason to return first, and keep offers for later steps and your highest-value customers.

Services behind this work

Lifecycle Marketing

From first tap to loyal fan, and back again when they drift.

AI Growth Solutions

AI where it earns its keep: fewer manual hours, better messages and proof against a control group.