You Don't Own the Room: The Real Math Behind Community-Led Growth
Photo: diverse team collaborating in bright modern office space with laptops and sticky notes, via statico.profootballnetwork.com
Every founder deck I've seen this fall has the same slide. A screenshot of a Slack workspace, a member count with a comma in it, and the phrase "community-led growth" set in 32-point type.
Then you ask the obvious follow-up — how many of those people showed up last week? — and the room gets quiet.
The 4% Everybody Quietly Accepts
Here's the pattern. A startup spins up a Discord or a Slack, invites the waitlist, ships a weekly newsletter, posts a launch thread. Six months later: 9,000 members, roughly 350 weekly actives, and a churn curve that hasn't moved a millimeter.
That's 3.9%. It's also entirely normal, which is the uncomfortable part. Most open chat communities settle somewhere between 2% and 8% weekly active participation, and the old 90-9-1 rule — lurkers, occasional contributors, regulars — still holds up better than most engagement dashboards. A number that big feels like traction. A number that small is the actual signal.
None of this means community is a bad investment. It means the slide is measuring the wrong thing, and the budget conversation that follows is built on sand.
Renting an Audience, Owning a Relationship
The cleanest test I know: if the platform disappeared tomorrow, what would you still have?
If the answer is "a member list I can't email," you rented an audience. Social followers, subreddit subscribers, Discord members — all of it lives on someone else's infrastructure, subject to someone else's algorithm, rate limits, and pricing page. A newsletter list is closer to ownership, but deliverability is still a landlord you don't control.
Owning a community means owning the relationships. That shows up as member-to-member threads, people answering questions you never asked, and a shared vocabulary that only exists inside your walls. It shows up as folks showing up for each other, not just for you.
What It Actually Costs
Community is the only growth channel where the tooling is nearly free and the labor is brutal.
A dedicated community lead runs $95K–$140K in most US markets, and that's before programming, events, swag, and the tooling stack. A functioning community of practice eats 15–25 hours a week of active programming — AMAs, office hours, moderation, escalation, member spotlights, the newsletter nobody wants to write. Budget the person, not the platform.
Run the per-member math and it looks like this: somewhere between $40 and $90 per monthly active member in year one, depending on how aggressively you program. Compare that to a $300–$800 blended CAC for SMB SaaS and it reads like a bargain.
But that comparison only holds if retention moves. If your community members churn at the same rate as everyone else, you haven't bought acquisition efficiency — you've bought a very expensive Slack theme.
Metrics That Predict Retention vs. Metrics That Look Good in a Board Deck
Predictive, if you're honest about them:
- Time to first response. If a new member posts and waits 36 hours for silence, they're gone. Median response time under four hours in the first week correlates with week-eight survival better than anything else I've seen.
- Member-to-member replies. Answers that come from other members, not staff. This is the single best proxy for whether you've built a community or a support queue with a nicer font.
- Week-four return rate. Of members who joined eight weeks ago, how many came back last week? Cohort it. Watch the slope.
- Share of questions answered by non-employees. If staff answer 80% of threads, you have a help desk. The crossover point — where members start carrying the load — is the moment community becomes an asset.
Vanity, in most cases:
- Total members. Always going up. Means nothing.
- Total messages. One power user in a bad mood can move this number 20%.
- Event RSVPs, impressions, follower counts, "engagement rate" scraped from a platform dashboard that defines it differently than you do.
- Anything that would look identical if you replaced your product with a different product.
The Cohort Test That Costs Almost Nothing
If you need to justify the budget, stop arguing about engagement rates and run the experiment.
Take your next two cohorts of new signups. Randomly invite half into a structured onboarding track with the community — a live welcome session, a named human, a specific ask in the first 72 hours. Leave the other half alone. Then measure net revenue retention at 90 and 180 days for both groups.
The trap here is selection bias: people who join communities are, by definition, the people who were already going to stick around. Randomized invitation is the only way through that. It's a boring methodology and it's the difference between a chart that survives diligence and a chart that doesn't.
A 90-Day Starter Sheet
- Pick one job. Community for support is a different product than community for practitioners. Pick one, or pick neither.
- Hire the human before the tool. A part-time community lead who shows up beats a $40K platform contract every time.
- Instrument first-response time and member-to-member replies. Two numbers. Put them on the wall.
- Kill two channels. Every additional surface splits the room. Consolidate.
- Write down what you'd still have if the platform vanished. That document is your actual community strategy.
The uncomfortable math isn't that community doesn't work. It's that community works slowly, expensively, and only when someone is doing the unglamorous relational labor. A Discord server is a room. A community is the reason people keep coming back to it.