The short version: three emails, entered at roughly 1.5 to 2 times the product's consumption cycle after the last order, with the diagnosis driving the content: beauty lapse is either substitution (they found another serum) or routine collapse (they stopped using the category), and the two need different arguments. The universal winback economics apply (2 to 5% of recipients converting per email is healthy, programs reactivate 12 to 20% of lapsed customers, and a save costs 5 to 10 times less than a new acquisition), and beauty brings its own weapon to the sequence: the mini-size re-trial, backed by the same sampling economics that power the checkout flow's ramp: over 35% of sample recipients go on to buy the full size.
Diagnosing the lapse: substitution or collapse
The supplement winback fights disbelief and the apparel winback fights drift; beauty lapse splits between two different exits. The substituted customer still runs the routine, just with someone else's product in your slot: the winback job is a switching argument (what changed, what they gave up). The collapsed customer stopped the routine entirely, usually in the week-two trough the post-purchase build exists to prevent: the winback job is restarting the habit, smaller and easier this time. You rarely know which exit a given profile took, so the sequence below argues both, in order of probability: substitution first (the bigger bucket for repeat buyers), collapse second (the bigger bucket for one-and-done customers). One-purchase lapsers and multi-purchase lapsers deserve different entry branches for exactly this reason.
Defining lapsed: the empties clock
- The threshold is consumption math, not a calendar default: 1.5 to 2 times the product's cycle since the last order, quantity-adjusted, per the same days-of-supply logic the replenishment flow runs. A 50ml daily moisturizer on an 8-week cycle makes day 90 a fair lapse line; a twice-weekly treatment makes day 90 mid-jar.
- Replenishment runs first, always: winback entry is replenishment-exit-plus-a-cycle, the standing sequence across every consumable build. The replenishment flow's already-reordered-elsewhere calibration tap is also this flow's best intelligence: those profiles are confirmed substitutions and can skip straight to the switching argument.
- Exclusions: active subscribers, open back-in-stock waiters (a customer waiting on your restock is loyal, not lapsed), and anyone inside the cancel-save cool-down.
The build: three emails, then the sunset handoff
Email 1: at entry. What changed since you left
The reopening leads with news, not nostalgia: the reformulation, the new size, the shade extension, the concern-line launch, merchandised to their concern property and purchase history. For the substituted, novelty is the only honest reason to switch back; for the collapsed, it is a reason to look again. Reviews accumulated since their last order carry the proof, filtered to their concern per the category's standing matched-proof rule. Zero guilt, zero discount.
Email 2: 5 to 7 days later. The routine restart
The collapse argument: the two-minute version of the routine, the single-product restart (not the five-step program that intimidated them out last time), and the honest consistency case borrowed from the skin-cycle logic: four weeks is when it shows. For confirmed repeat buyers this email leans switching instead: the ingredient-transparency comparison, the concentration honesty, what the formula does that the drugstore substitute does not, kept in cosmetic-claim territory as always.
Email 3: 5 to 7 days later. The re-trial ramp, then stop
Beauty's native final offer is not a percentage, it is the mini: the trial size of what they used to buy (or its reformulated successor) as the low-risk re-entry, with the full-size path and subscription option beside it. The sampling data makes this the margin-safe close: a third-plus of trial recipients convert to full size, which beats what a 15%-off code buys in a category where discount-trained buyers churn hardest. Where no mini exists, one capped offer below welcome-level with a real expiry, per the universal offer discipline. Then the flow ends and non-responders route toward the sunset track.
Routing notes
- One-purchase versus repeat at entry: the one-and-done lapser gets the collapse-weighted track (restart small); the multi-purchase lapser gets the substitution-weighted track (what changed, what compares).
- The concern property ages: skin changes with seasons and years. Email one can refresh it with the one-tap concern question, which doubles as an engagement signal that re-sorts the branch.
- Shade fidelity holds: complexion-product winback references their exact shade family or offers the two-question shade re-check, never a generic product shot, per the match-anxiety rules.
- Reactivated customers route into post-purchase as if new: the routine onboarding is precisely what the collapsed lapser needed the first time.
What to measure
- Per-email conversion against the 2 to 5% band and program reactivation against 12 to 20%, measured across the sequence plus 30 days
- Mini-versus-full-size share of reactivations, judging the mini cohort on its 60-day upgrade rate per the sampling economics, not first-order AOV
- Track performance by entry branch: substitution-weighted versus collapse-weighted conversion, which tunes the diagnosis split over time
- Second-cycle survival of reactivated customers: a winback that produces one order and a second lapse was a coupon; route-through to replenishment is the honest success metric
- Subscription share of reactivations, the number that says the habit rebuilt rather than rented
Frequently asked questions
When should a beauty winback flow trigger?
At 1.5 to 2 times the product's consumption cycle since the last order, quantity-adjusted, after the replenishment flow has run and missed. Flat 90-day triggers are wrong in both directions across a catalog with 30-to-90-day cycles.
What makes beauty lapse different?
It splits between substitution (another product took your slot in a routine that continues) and collapse (the routine stopped). The two need different arguments, and purchase count is the best available predictor of which happened.
Should beauty winback lead with a discount?
No. Lead with what changed and the routine restart; close with the mini-size re-trial, which converts a third-plus of recipients to full size at intact price integrity. The capped-discount fallback exists only where no trial size does.
What does the winback owe the replenishment flow?
Its entry logic and its intelligence: winback starts where replenishment failed, and the calibration tap's already-reordered-elsewhere answers pre-sort the substitution branch.
Where do reactivated customers go?
Into post-purchase as if new: the routine onboarding, the check-in, the results-window review ask. The collapsed lapser churned for lack of exactly that arc the first time.
Sources
- Eightx. Win-back and reactivation rate benchmarks for DTC.
- Free Yourself. Beauty product sampling conversion rates.
- 303 London. Premium skincare email benchmarks.