An online bookstore was losing regular readers silently — noticing months after they'd drifted away. We built an engine that learns each customer's buying rhythm and triggers a tiered win-back journey the week they lapse. 18% of lapsed buyers now come back.
Typical results
7 days
To spot a lapse, down from 90
58%
Win-backs with no discount
100%
Journeys stop on purchase
Figures reflect typical results for this kind of automation.
The challenge
Every customer buys on their own rhythm — some monthly, some twice a year — so a fixed '90 days inactive' segment missed both kinds. Lapses were spotted around 90 days late on average, win-backs were generic coupon blasts reactivating barely 4% of recipients, and the discounts trained loyal readers to simply wait for the next one.
Before and after
Lapsed buyers reactivated
4.5×Days to spot a lapsed customer
−92%Win-backs requiring a discount
−58%The automation
We built a purchase-cycle engine on the store's order history: it computes each customer's personal repurchase interval and flags anyone drifting 50% past it. A tiered journey then runs automatically — a personal email of new releases in their favorite genres first, a WhatsApp nudge after ten days, and a time-limited discount code only on day 21. Any purchase exits the customer instantly, and every journey is logged to the CRM.
How it works
- 01
Learn each cycle
The engine studies every customer's order history and computes their personal repurchase rhythm — 3 weeks for some readers, 3 months for others.
- 02
Spot the drift
A nightly Make scenario flags anyone drifting 50% past their own cycle — within 7 days of it happening, not at quarter's end.
- 03
Content first
The first touch is a personal email of new releases in the genres they actually buy. No coupon, no pressure.
- 04
Incentive later
Still quiet after ten days? A WhatsApp nudge follows, and only on day 21 does a time-limited discount code appear.
- 05
Stop on purchase
Any order exits the customer from the journey instantly and resets their cycle clock.
Starts from
- Order history
- Live store orders
Delivers
- Personalized win-back emails
- WhatsApp nudges
- Timed incentive codes
The impact
Lapsed customers are now spotted within 7 days instead of 90, and 18% of them reactivate — up from 4% with the old coupon blasts. Better still, 58% of win-backs needed no discount at all, and journeys stop the moment a customer buys, protecting margin and goodwill alike.
Questions about this automation
How does the automation decide that a customer has lapsed?
It looks at each customer's own order history and works out their typical gap between purchases. A nightly job flags anyone who is 50% past that personal interval. A monthly buyer is flagged after about six weeks and a twice-a-year buyer only after several months, so neither group is missed or nagged too early.
Why does the journey start without a discount?
The first touch is a personal email with new items in the categories the customer actually buys. A WhatsApp nudge follows after ten days, and a time-limited code appears only on day 21. Many customers return on the content alone, which protects margin and avoids training loyal buyers to wait for coupons.
What happens when a customer buys during the journey?
Any new order removes the customer from the journey at once and resets their cycle clock, so no further reminders or codes go out. Each step and its outcome is logged to the CRM. Simple automations like this are usually live within days, and the timing and thresholds are then tuned on real order data.