The Membership House

Book a call

Inside a six-figure community's churn problem

How the House's founder found the leaks, and built the system that catches members before they leave.

A paid community with hundreds of members and six figures a year in recurring revenue. Growing steadily on the front end through content and ads, but losing members every month out the back at the industry-typical rate of 6 to 9%. That's the part worth sitting with: this was a strong community. Great content, a founder members genuinely liked, healthy growth. And it still leaked, because almost every community does, at the same predictable moments. At 9%, a community replaces nearly half its member base every five months just to stand still.

The founder was excellent at getting people in. Nobody owned keeping them. That's not a failure. It's the industry default, and it's fixable.

I went through the churned and cancelled member list one by one. Every exit got a personal DM asking one honest question: what made you decide not to continue? Then I mapped every answer and every exit against tenure.

The pattern was not random. Exits clustered into distinct lifecycle windows:

• Week one: the biggest leak. Members who joined, were active for a few days, then went silent and never came back. One member was active her entire first week, then disappeared and cancelled at day 38. She never had a first win. By day 8 she was already halfway out the door.
• Months 3 to 4: the ROI evaluation. Members deciding whether the membership was paying for itself.
• Renewal moments: annual members quietly not renewing, and long-tenure monthly members re-evaluating on founding rates.
• Payment declines: members who never chose to leave at all. A card failed and nobody told them.

Underneath all of it: an onboarding survey of ~200 members showed 47% had never posted or spoken. Silent members do not complain before leaving. They just leave.

Instead of reacting to cancellations, I built a retention operating system that covers every exit path.

Four messages, each firing at a specific moment:

• The save: within 24 hours of a cancellation, while access remains. Warm, direct, human.
• The check-in: proactive, fired at 21+ days of silence and before month 3, 6, and 11 renewal moments. Catches the decision before it is made.
• The payment recovery: same-day, removes the billing barrier for members who wanted to stay.
• The feedback ask: 7 to 14 days after full churn, when answers get honest.

Seven upstream goals so those messages fire less often: week-1 activation (target: 60% of new members posting by day 7), lurker-to-contributor engagement, monthly-to-annual ascension, member wins surfaced weekly, referral cultivation, feedback loops before cancellation, and a 90-day re-entry path.

The machine underneath: a churn watchlist and community health dashboard, AI-powered analysis that flagged at-risk members weekly, automated call-transcript mining for member pain points, and a weekly operating rhythm. The founder got a pattern report every week instead of anecdotes.

• Churned members responded. Exit messages got real answers: timing, travel, unmet expectations. Reasons became patterns, patterns became fixes.
• Cancelled members were saved and re-entry candidates identified, including a trial member who asked to come back.
• Payment-decline members were recovered who would otherwise have churned without ever deciding to.
• The founder went from guessing about churn to a weekly report naming exactly who was at risk and why.
• The system outlived my hours: the analysis, drafting, and reporting run in minutes, not days.

“What stood out most was her ability to move between the strategic and the operational without losing quality on either end. I'd recommend Aqeelah to anyone looking for someone who can build, think, and deliver.”

Most community owners treat churn as a mystery or a mood. It is neither. It is a set of predictable leaks at predictable lifecycle moments, and each leak has a specific fix. I find the leaks in one week. Installing the fixes takes three or four.

Book a Call