Win-back flows: 12 examples that actually work

The most common failure in a win-back program is using "hasn't opened in 90 days" as the trigger. Opens are a weak signal — Apple Mail Privacy Protection inflates them for over half the market — and 90 days is arbitrary.

Better triggers use a real behavioural signal. For a subscription product: no login in N days, where N is tuned to your natural usage cadence. For a marketplace: no purchase in N days plus no session start in M days. For a content product: no content consumption event in N days plus no app open in M days. The principle: lapse should reflect the value the user gets from the product, not just whether they read a marketing email.

And the threshold should reflect natural cycle time. A food delivery product reasonably defines lapse as 14 days since last order; an annual travel booking product needs 9+ months. Pick based on the natural cadence of the product, not a round number.

Lapsed users aren't one audience; they're at least three. Tiering them is how you avoid sending a "please come back" email to someone who cancelled last week because of a billing dispute.

Tier 1 — Recently lapsed, high historical engagement. These are the users most likely to come back. Light touch, high-value reminder, short sequence (2–3 messages).
Tier 2 — Medium-term lapsed, moderate historical engagement. Needs a real reason to return — a product update, new feature, changed offering. Slower sequence (3–5 messages over 4–6 weeks).
Tier 3 — Long-lapsed or cancelled. Typically the largest audience and the lowest conversion rate. Most of these users should eventually be suppressed; the win-back attempt is a final qualification before sunset.

The Orbit Win-Back Playbook skill generates the tier definitions, the sequence per tier, and the sunset policy — tuned to your specific lapse definition and product cadence.

1. The "here's what you missed" digest.Summarise 3–5 things the user missed: new features, popular content, community milestones. Works for content-heavy and community products. Avoid for transactional products — you'll look out of touch.

2. The "what changed since you left" update. Single most impactful product change since the user lapsed. Works particularly well if the change addresses a common complaint. Lead with the change, not the ask to return.

3. The personalised usage memory."You last listened to X", "Your favourite item is back in stock", "Your last workout was Y weeks ago". Pulls a specific past engagement data point into the message. Requires a real data model behind it; the generic version reads as creepy rather than personal.

4. The low-friction re-entry. One-click resume, one-click restart. Removes the friction of reconfiguring preferences or remembering passwords. The surprising learning here is that removing password friction alone can lift reactivation 2–3x among tier 1.

5. The direct discount offer. The pattern most teams default to. Works for tier 2 and 3, but the discount level matters: 10% is usually not enough to move someone who stopped caring; 30% signals desperation and trains users to wait. 15–25% is the typical sweet spot, tied to a specific product or category rather than blanket.

6. The "what went wrong" survey.A single open-ended question asking what caused the lapse. Low conversion to reactivation but high value as a product-research input. Use as a late-sequence message in tier 3 — the users who engage with the survey are giving you honest feedback that's hard to get anywhere else.

7. The behaviour-based trigger. Not a campaign — an automation. If a lapsed user takes any small action (opens an email, clicks a link, visits a page), trigger a targeted flow that day. Works because intent-in-the-moment is a much stronger signal than the lapse itself.

8. The peer-behaviour nudge."Users like you are now using X." Social proof for re-engagement. Requires a real user model behind the recommendation; generic "popular with other users" reads as weak.

9. The "we'll miss you" pattern.A direct-from-the-team note, signed by a real person, brief, no discount. Often run as the final attempt before sunset. Works because it breaks the pattern of every other marketing email the user has received, and because the implicit message is "this might be the last one, so respond now or never".

10. The preference-change invitation."Too many emails? Tell us what you'd prefer." Preserves the subscriber relationship without forcing reactivation. A subscriber who opts down in frequency is worth vastly more to your deliverability than one who marks you as spam because you kept sending.

11. The seasonal tie-in. Align the win-back moment with a natural re-evaluation point: New Year, fiscal quarter end, seasonal product relevance, annual usage anniversary. Higher inherent relevance makes the ask feel less forced.

12. The channel-switch. Move lapsed email subscribers to a lighter-weight channel — push, SMS, or in-app only — rather than forcing continued email receipt. Reduces complaint rate, preserves the subscription relationship, and often re-engages users who were tired of the email cadence specifically. Requires the user to have opted into the secondary channel, which is often the constraint.

Every win-back program needs an explicit sunset policy — the rule that decides when a user stops getting emails entirely. Without it, your lapsed list grows indefinitely and drags your sender reputation down over time, because every send to a long-lapsed user is a low-engagement data point ISPs use to judge you.

A typical policy: after the win-back sequence completes, if the user has not taken any engagement action (open, click, login, purchase, session start), remove them from marketing sends. Transactional sends continue. Re-opt-in is available via a link in a final "goodbye" email or via a product surface.

The hard number that matters: list churn rate. Healthy programs sunset 1–3% of their list each month. Programs that sunset nothing end up sending to a list that's 40% dead weight within 18 months. The deliverability guide covers how list hygiene compounds into sender reputation, and the Deliverability Management skill runs the full diagnostic.

The primary metric for a win-back program is reactivation rate: % of the sequence audience that took a defined reactivation action within the measurement window. Define the action specifically: "completed a purchase" is different from "opened any email". Measure both, but the purchase metric is the one that matters for ROI.

Secondary metrics: complaint rate (should stay under 0.1%; winback programs have higher natural complaint rates because the audience is ambivalent), unsubscribe rate (higher than your average campaign is normal and actually healthy — it's the sunset working), and downstream LTV of reactivated users (do they stick, or churn again within 30 days?).

The trap: measuring the open or click rate of the winback sequence itself. Open rates are 30–40% of your baseline for this audience and will look terrible. Use reactivation as the primary metric and judge the program on that number, not on the sequence engagement in isolation.