All use cases

Win back

Lapsed-tracker win-back with progress nostalgia

Their 30-day streak and the goal they set are still sitting there — so instead of a discount they never asked for, the agent shows them their own progress and makes picking it back up one tap.

Get aheadTime-based triggerBehavioral triggerPushEmailWhatsAppSMSIn-appWellness (B2C apps)Subscription

What it does

Daily-tracker wellness apps lose members quietly: a fitness, cycle, nutrition, or meditation tracker goes dark for weeks, and the standard win-back is a blanket discount that ignores everything the member built. But a lapsed tracker user is not a cold lead — their streak, their logged history, and the goal they set are all still on the account, and that personal progress is a far stronger hook than money off. The agent reactivates with the member's own history: it recalls their best streak and the goal they were chasing, and offers a low-effort, one-tap way to start again tonight. It is cheap, history-personalized re-engagement that works whether the member is still subscribed but dormant or has churned outright, and it reads as encouragement rather than a sales push.

How it works

  1. 1
    Trigger. a daily-tracker member crosses a dormancy threshold — no logging for weeks — or has churned, opening a per-member reactivation window rather than a single blast to the whole lapsed list.
  2. 2
    Decision. the agent reads the member's saved progress (best streak, logged milestones, the goal they set, what they used most), picks the nostalgia hook and the lowest-friction restart for this person, and a judge clears the drafted message for tone and relevance before it sends.
  3. 3
    Action. a warm, progress-first message lands on the member's preferred channel, naming their own milestone and offering a one-tap restart that pre-fills tonight's entry — not a discount and not a pricing page, just the easiest possible step back in.
  4. 4
    Follow-through. if the member returns on their own first, pending outreach is cancelled; the outcome writes back to memory; and reactivation is measured per cohort against a randomized holdout so real lift is separated from members who would have come back anyway.
Example
AgentHey — it's been a few weeks, and I didn't want your progress to slip away. You held a 30-day logging streak in March and you were close to the goal you set yourself. All of it is still saved and waiting. Want to pick it back up tonight? One tap logs today and you're going again.
Customerhonestly I just fell off and felt behind
AgentThat's the most normal thing in the world — falling off doesn't erase what you built, and your history's all still here. No restart from zero. Tap once and tonight just continues the streak you already earned.

Configuration

How the agent is wired for this use case.

Triggera tracker_dormant event (a regular logger past a weeks-long no-activity threshold) or a subscription_lapsed event, opening a per-member reactivation window from the app backend.
Tools & actions
  • App backend · read logging history, best streak, milestones, and the member's set goal to build the nostalgia hook and pre-fill the one-tap restart.
  • Member memory · recall what the member used most, the reason for lapse if captured, and channel preference; record the win-back outcome.
  • Messaging channel · deliver the progress-first reactivation message on the member's preferred channel and run the conversation.
  • App restart action · pre-fill the member's usual entry so resuming the streak is a single tap, with no re-onboarding.
  • Holdout ledger · assign the member to test or control and record reactivation per cohort.
Autonomydormancy detection, the progress-recall message, and the one-tap restart run unattended under judge gating, quiet hours, and frequency caps; nothing money-moving is involved — this is a habit restart, not a billing action, and any re-subscribe payment for a churned member is left to the client's secure checkout step; outreach is suppressed the moment the member returns on their own.
Channelspush · email · whatsapp · sms · in-app
Escalationa do-not-contact or data-deletion request, or a reply that needs a billing or account action beyond the restart, hands off to the client's support route with context; a member who opts out of win-back messages is excluded.

What you need

The inputs this use case runs on. Your channels stay yours; the agent supplies the judgment.

Signals

logging events and the dormancy threshold, subscription-lapse events, organic-return events for stale-send cancellation, and member replies.

Data

per-member logging history, best streak, milestones and the set goal, captured lapse reason where available, channel consent and contact tokens, persistent memory of the member's pattern.

Guardrails

fires on genuine dormancy, not on a single missed day; judge gating on every unprompted message with "don't send" as a valid outcome; suppression of recently contacted, opted-out, or already-returned members; frequency caps and quiet hours; a progress hook rather than a discount by default; any re-subscribe payment via secure hand-off, never in-thread; per-cohort holdout measurement; the client keeps owning the send channels.

Metrics it moves

  • reactivation-rateup: a member's own streak and goal pull them back more reliably than a generic discount they never asked for.
  • ltvup: a one-tap return to a saved habit revives a relationship that a blanket win-back would have let lapse for good.
  • open-rateup: a subject line built from the member's real progress reads as personal encouragement, not marketing, so more of them open it.

Related use cases

Win back

Lapsed-subscriber reactivation

the general subscription win-back this specializes with a streak-nostalgia hook and a one-tap habit restart

Win back

Memory-personalized win-back

the all-industry one-to-one win-back motion this draws on

Retain / Save

Logging-streak rescue save

the before-the-break sibling that saves the streak rather than reviving it after dormancy

See it on your own customer journey

Bring one drop-off, one churn cliff, or one silent segment. We will show you what a proactive agent with memory and judgment does with it.

Book a demo