The problem this solves
Every list quietly divides into the living and the dormant, and the dormant fraction only grows. Subscribers who have not opened in a quarter are past the reach of ordinary campaigns, yet they keep receiving them, which is the worst of both worlds: no revenue from the sends, and a slow accumulation of engagement damage as mailbox providers watch message after message go ignored. A list that is one-third dormant is not a bigger list. It is a smaller list dragging an anchor.
Teams avoid the win-back project for an understandable reason: its honest ending is unsubscribing people, and nobody enjoys building the machine that shrinks their own headline metric. So the dormant segment sits unaddressed while the sender reputation absorbs the cost, and the eventual reckoning - a deliverability problem that finally forces list hygiene - is far more expensive than the graceful goodbye would have been.
A proper win-back series holds two goals in tension: a real attempt at rekindling, and a clean exit for those who will not return. Re-permission asks the honest question, best-of leads with the strongest material rather than an apology, last-chance makes the stakes plain, and sunset logic removes non-responders on schedule. The structure is standard; what stops teams is drafting it well and wiring the exit logic correctly. Both are buildable.
How the mission runs
- The dormant segment is defined precisely. The agent specifies the entry condition with you - no opens or clicks in 90 or more days, with any exclusions you need, such as recent purchasers whose engagement lives outside email. A precise definition matters because everything downstream, including the sunset, keys off membership in this segment.
- Three emails are drafted with distinct jobs. Re-permission asks plainly whether they still want to hear from you, making staying one click. Best-of leads with your strongest recent material, arguing by demonstration that the list is worth their attention. Last-chance states honestly that this is the final email unless they act, respectful rather than dramatic.
- The sunset logic is written out explicitly. The exit rules are specified in plain language before the build: what counts as re-engagement at each step, how re-engaged subscribers leave the series and rejoin normal flows, and what happens to non-responders after the last email - suppression or removal, per the policy you choose.
- You approve the copy and the exit policy together. The series and the sunset rules come to you as one reviewable package, because the emails only make sense against the exits they promise. The last-chance email in particular must mean what it says. Your sign-off covers both the words and the consequences before anything is built.
- The series is loaded into Omnisend, staged. On approval, the agent builds the automation in Omnisend: segment entry, the three emails with agreed spacing, re-engagement exits at each step, and the sunset action at the end. It reports exactly what was created, staged for your review, so the series starts only when you activate it.
The prompt
This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.
What comes back
A complete win-back automation staged in Omnisend: a precisely defined 90-day dormancy segment, three emails with distinct jobs and honest copy, re-engagement exits that return revived subscribers to your normal flows, and sunset logic that suppresses or removes non-responders on the schedule you approved. The written exit policy ships alongside, so the series' promises and consequences are documented rather than folded silently into automation settings.
Make it yours
- Add an incentive test: half the dormant segment sees a best-of email, half sees an offer, and the re-engagement gap tells you whether discounts actually revive your list.
- Stagger by dormancy depth: 90-day dormants get the full three-email attempt, while 180-day dormants get a single respectful re-permission before sunset.
- Precede the send with a deliverability review of the dormant segment, so the win-back itself does not spike bounces from addresses that died during the silence.
Frequently asked questions
Will this actually unsubscribe people from my list?
Only per the exit policy you explicitly approve, and only after the full series has run its course without re-engagement. Suppression rather than deletion is the default recommendation, which stops sends while preserving the record. The approval covers the policy in writing, so nothing about the sunset is implicit.
Is emailing dormant subscribers a deliverability risk in itself?
Mailing deeply disengaged addresses carries some risk, which is why the series is short, spaced, and permission-framed rather than promotional. The staged review is the moment to align it with your sender practices, and pairing it with a deliverability check on the segment is a supported variation for cautious senders.
What does the agent change in Omnisend, and can I undo it?
It creates the segment definition and the win-back automation, staged and inactive, and alters none of your existing flows or campaigns. The completion report lists every created object. Undoing before activation is deleting the staged automation; after activation, the exits and sunset only ever act on the approved policy.