A win back campaign is the timed sequence you send to customers who cancelled a subscription or stopped reordering, and its only job is to get them back onto a recurring order. In health and wellness DTC, a segmented one recovers 8% to 15% of lapsed subscribers within 90 days. A third to a half of that recovery usually comes from fixing failed payments, which is the part nobody builds first.
Most winback email sequences underperform because every lapsed subscriber gets the same message and that message opens with a discount. A coupon in email one teaches your list that cancelling produces a coupon, and it pulls back the customers with the lowest lifetime value while the people who left for a solvable reason never hear the answer to the problem they actually had.
I’ve worked the retention side of health and wellness brands from a few million up through Ancestral Supplements at $28M to $60M+ and Dr. Gabrielle Lyon’s $6M practice. Every number on this page is my operator range for that category. Yours will shift with price point and cadence.
Segment before you send
Five lapse reasons are worth separating in supplements, because four of them get different answers:
- Still have product. Your ship cadence is faster than their consumption. Very common with twice-daily capsule products shipped every 30 days. These people like you.
- Didn’t feel results yet. They quit inside the window where the product hadn’t done anything visible. An expectation problem you created at checkout.
- Price. Real, and smaller than founders assume.
- Payment failed. Involuntary. Never a marketing problem, and it lands in your churn number anyway.
- Switched brands. They’re taking someone else’s version now.
Reasons one and two are the biggest slice in this category, and both are answerable with information. Nobody needs 20% off to understand that they should shift to every 60 days, or that the thing they were waiting for shows up in week three.
How to capture the reason: a single required question on the cancellation screen, radio buttons, five to six options written in customer language, no “other” as a first-class choice. One click, then the cancellation proceeds.
A free-text box collects nothing you can use. Most people skip it, the rest type “no longer needed,” and none of it routes into a flow automatically. You want a stored field on the subscription record that a sequence can branch on.
The sequence, send by send
Six sends across 90 days. The discount lives at send five.
| Send | Timing | Job | Offer | Expect |
|---|---|---|---|---|
| 1 | Day 3 | Acknowledge the cancel and answer the stated reason directly | None | 38-48% open, 5-8% click, ~15% of recovery |
| 2 | Day 10 | Teach the mechanism or the timeline they missed | None | 30-40% open, 4-6% click, ~20% of recovery |
| 3 | Day 21 | Offer a structural restart: skip, pause, slower cadence, smaller size | Structural only | 28-36% open, 4-7% click, ~25% of recovery |
| 4 | Day 38 | Show what’s changed and what other customers report | None or free gift with restart | 24-32% open, 3-5% click, ~10% of recovery |
| 5 | Day 55 | The offer, with a deadline | 15-25% off first restart order | 22-30% open, 4-8% click, ~25% of recovery |
| 6 | Day 85 | Last send, plus a preference or suppression choice | None | 18-26% open, 2-4% click, ~5% of recovery |
Three quarters of the recovery happens before the discount arrives, which is the whole argument for this ordering.
What each email has to do
Send 1, day 3. Branch on reason. To the “still have product” group, name it out loud and offer a cadence change in one tap. To the expectation-gap group, give the real timeline and what to watch for. Subject lines run plain and specific: “About that half-full bottle” or “You’ve probably got extra on the shelf.”
Send 2, day 10. This is the education email and it carries more weight than anything else in the flow. Explain the mechanism and why the first few weeks feel like nothing. Cite what your own customers report at week four and week eight. Subject lines: “Week 3 is usually when it shows up” or “What most people notice first.”
Send 3, day 21. The structural restart. Every control you have that isn’t a discount goes here: skip a shipment, pause 60 days, ship every 8 weeks instead of every 4, drop to the smaller size. Subject lines: “Ship it every other month instead” or “Want it slower?”
Send 4, day 38. Reformulations, new third-party testing, a new flavor, a bundle that fits their routine. If nothing changed, use customer stories from people who restarted. Subject line: “What we changed since spring.”
Send 5, day 55. Now the offer. Make it one time only, with a real deadline, and apply it to a subscription restart rather than a one-off purchase. Subject lines: “20% off your restart, good through Sunday” or “Your restart discount expires Friday.”
Send 6, day 85. Short. Tell them this is the last one, and give them an “email me twice a year” option instead of a hard unsubscribe. You’ll keep more addresses that way.
Put a reply-to on every send that a human actually reads. Replies to send two are the best qualitative retention data you’ll get all quarter.
Failed payments deserve their own build
Involuntary churn is commonly 20% to 40% of total subscription cancellations in this category, and it’s the highest-ROI fix on this page because those customers never chose to leave.
Retry schedule: attempt at failure, again at 48 hours, again at day 5, again at day 9, final attempt at day 14, then cancel.
Spacing matters more than the count. Four attempts inside 72 hours hit the same empty balance four times and burn your processor’s risk score. Spread across two weeks, you catch the paycheck.
Around that schedule, the messaging cadence:
- Hour 1: email, subject line states the problem plainly, one button to a hosted card-update page. No login required.
- Day 3: second email, different subject, same link.
- Day 6: SMS. This is the single highest-converting message in the whole sequence. One sentence and a link.
- Day 12: final notice email that names the cancellation date.
Turn on your processor’s card account updater so expired and reissued cards refresh automatically before they ever fail. Most subscription platforms support it and most brands leave it off.
With that setup, expect 55% to 75% of failed payments to recover. Processor defaults with no dunning sequence land closer to 30% to 40%. That gap is real money on a $6M brand, and the build takes about a week. It belongs near the top of any list of what to automate first.
The pre-lapse save beats the win-back
A good cancellation flow saves 20% to 35% of attempted cancellations. That’s three to five times what the win-back sequence recovers, on the same population, before they ever leave.
Offer these in order, and put the discount last:
- Skip the next shipment. Solves “I still have product” at zero margin cost and no change to the plan.
- Pause for 30, 60, or 90 days. Solves a temporary reason. Keeps the payment method and the price they signed up at.
- Change the cadence. The permanent version of the same fix. A subscriber on 8 weeks is worth far more than a cancelled one on 4.
- Downgrade. Smaller size or a slimmer bundle. Answers price without publishing a lower price to your whole list.
- Discount. Last, and time-boxed. Once you show it here, a share of your subscribers will find the cancel button on purpose.
Brands hide these controls to protect a churn number and it backfires every time. A customer who can’t find “skip” cancels instead. Same logic as the replenishment work in customer loyalty campaigns for supplement brands.
The math
Win-back recovery rate: subscribers reactivated within 90 days ÷ subscribers who lapsed in that cohort
Measure it against the lapsed population. Send-based reactivation rates flatter the flow and hide the size of the hole in the bucket.
Incremental margin: (reactivated subscribers × expected post-restart LTV × contribution margin %) − discount cost − sequence build and send cost
Discount cost against recovered LTV: discount dollars given ÷ contribution margin earned from those reactivated customers
Run that third one separately for the discount cohort. If a 25% off restart pulls back customers who churn again in one cycle, you paid to move revenue forward and lost margin doing it. That’s the failure mode of discount-first win-back, and it hides inside a healthy-looking recovery rate.
Benchmarks worth holding yourself to
- Total win-back recovery: 8% to 15% of lapsed subscribers within 90 days, all sources combined.
- Marketing-driven share: 4% to 8% of the lapsed population. The rest comes from dunning.
- Failed-payment recovery: 55% to 75% with a real retry schedule and card updater.
- Cancellation flow save rate: 20% to 35%.
- Timing window: recovery falls off a cliff after 120 days. Past six months, treat them as a cold list to reacquire.
For the surrounding numbers, see subscription retention benchmarks for supplements.
Who to let go
Not every lapsed subscriber is worth a sequence, and chasing the wrong ones costs margin and deliverability.
Let go of anyone who cancelled after a single discounted order. They came in on a coupon and they’ll leave on the next one. Their post-restart LTV rarely clears the cost of the discount that brings them back.
Let go of anyone who reported that the product disagreed with them. Send one honest email offering a different SKU, then stop. Pushing a product that made someone feel bad is how you buy a refund and a review.
Let go of chargeback and serial-refund accounts. Suppress them at the subscription level.
Let go of anyone with zero opens across all six sends. Keeping them in the sendable file drags your engagement rates down and pushes the rest of your list toward the promotions tab. Suppress at send six and check back once a year.
A tighter list with a 40% open rate recovers more subscribers than a bloated one at 18%.
Common questions
How long should a win-back sequence be? Five to six sends across 60 to 90 days. Recovery drops sharply after day 120, so a longer sequence mostly adds unsubscribes. Front-load the education sends and put any offer past day 50.
Should I offer a discount to win someone back? Yes, at the end. A 15% to 25% one-time restart offer at send five converts the price-sensitive segment without teaching the rest of your list that cancelling produces a coupon. Lead with it and you recover your cheapest customers and train the behavior.
What’s a good win-back recovery rate? 8% to 15% of lapsed subscribers within 90 days in health and wellness DTC. Roughly half of that comes from failed-payment recovery, so if your dunning is weak, expect the low end and fix dunning first.
When is a lapsed customer gone for good? Around 120 days post-lapse the response rate collapses. Past six months, stop treating them as a win-back audience and move them into standard reacquisition through ads and broadcast email. The economics change and so should the messaging.
How do I reduce failed-payment churn? Turn on card account updater, retry at failure, 48 hours, day 5, day 9, and day 14, and wrap that with emails at hour one, day three, and day twelve plus an SMS on day six. That setup moves recovery from the 30% to 40% range up to 55% to 75%.
If you’re a wellness brand losing subscribers faster than you’re replacing them and the win-back flow is a single “we miss you” email with a coupon, there’s a month of work here that pays for itself in the first quarter. Here’s how engagements work, and the button below books thirty minutes.