Lifecycle ops — onboarding, retention, expansion, win-back — handled by an agent that reads your event stream, files a morning briefing, and acts on what you approve. Not another dashboard. Not no-code. A colleague.
The problem
"The hard part of building a SaaS isn't the SaaS. It's the thousand quiet conversations with users that never happen because the team is small."
Your event stream knows when activation drops, when a power user goes cold, when a feature finally clicks. You don't — because you're shipping. Patterns rot in the queue.
Mixpanel shows you a dip. Customer.io will send the email. But nothing connects the noticing to the doing — and nothing writes the campaign for you with the right tone for that specific cohort.
Onboarding, win-back, expansion, NPS, weekly digest — at a real company this is 2–3 people. You're one. So most of it doesn't happen. So most of your TAM never sticks.
What ChurnLoop does
ChurnLoop is not a chatbot you query. It runs in the background against your event stream and your tools, and pulls you in only when there's something worth your attention or a decision worth your call.
Reads every event, every funnel, every user reflection. Spots the things that moved — not just the metrics that fell, but the cohorts where they fell, and what those users have in common.
Every morning, a briefing in your voice: what happened, what it probably means, what you might want to do about it. Written, not dashboarded. Read it in two minutes with coffee.
Drafts the email, picks the audience, writes the in-app prompt, schedules the campaign. Hits send when you tap approve — or auto-runs the playbooks you've marked safe.
A day with ChurnLoop · Monday, May 11
While you slept, ChurnLoop ran overnight analysis on 14,200 events from 893 users. Two trend changes worth knowing about, one quiet win, three actions awaiting your call. Two-minute read.
893 users analyzed · 14.2k eventsAudience built. Email drafted in your voice (the one ChurnLoop learned from your last 40 sends). Personalization tokens filled. Sent to 4 stuck power users and 12 ceiling-pushers. You snoozed the routine one for tomorrow.
16 users contacted · 2 campaigns liveA playbook you marked safe ran without you. ChurnLoop sent the team lead a soft heads-up about the upgrade path, CC'd the workspace owner, and pinged your Slack with a one-line summary. No interruption.
Playbook: ceiling-warning · autopilotLogin frequency down 80% across two weeks. Feature usage decayed. Last support ticket marked "frustrated." ChurnLoop paused all automated outreach to her, drafted a personal note from you, and surfaced it on the briefing for tomorrow.
Risk score: 0.78 · auto-pause activeA short PM digest: 23 outbound messages sent, 11 replies received (8 positive, 3 frustrated — ChurnLoop already flagged them). Two playbooks proposed for tomorrow's briefing. One funnel regression worth a deeper look.
Replies routed · summary < 90 secondsThe surfaces
Nothing is a dashboard for the sake of being one. Each surface answers a specific question a founder asks themselves.
Manifesto
We made a list of every category ChurnLoop could be mistaken for. Then we drew a clear line away from each of them.
Trust & control
ChurnLoop defaults to copilot — drafts, then asks. Autopilot is opt-in, per playbook, with hard caps. You can read the trail of every decision ChurnLoop ever made on your behalf.
Every playbook has a mode toggle. Some you'll always want to read first (win-back from you personally). Others should just run (welcome emails on Sunday).
No more than 2 messages per user per week. Quiet hours respected per timezone. Every email includes a one-click unsubscribe — opted-out users are never contacted again.
For each message, each pause, each escalation: the events ChurnLoop saw, the reasoning it followed, the playbook it ran. You can rewind any user's history.
DSR requests are one click — ChurnLoop cascades erasure across all stored events, reflections, and dispatches. Export your raw data to S3 or a Postgres mirror anytime.
Your event stream, your users, your sent emails — none of it trains the underlying models. Bring-your-own-key for completions if you prefer.
A red switch in the topbar pauses every playbook, every campaign, every autopilot run. Your audit trail stays. You decide when to resume.
Pricing
Every plan includes every surface, every playbook, every model call, and every seat. The only thing that scales is how much of your event stream ChurnLoop is watching.
For early-stage products shipping to real customers for the first time.
For growing SaaS with an active user base and a churn problem worth solving.
For teams operating at scale or running multiple products.
Questions you probably have