Most “communities” launched by startups die within a year. Not because community-led growth doesn’t work, but because of how it was launched: as a marketing channel, anchored in a growth plan, staffed by those who had time on their hands, and measured by a metric that no one could tie to revenue. A Slack workspace with 400 members and three messages per week is not a growth strategy. It’s a ghost town with a cute logo.
Companies that make community a true engine of growth are doing something different, and it shows in the way they treat it. They build the community as a product, with its own roadmap, its own owner, and its own reason for existing beyond funneling people to a sign-up button.
We earn a commission if you make a purchase, at no extra cost to you.
We earn a commission if you make a purchase, at no extra cost to you.
The reveal: is it run as a product or a side project?
Look at what the winners actually invested. When Figma filed for an IPO, its community wasn’t a footnote in the marketing section; he was described in the S-1 as a platform moat, with more than 250,000 community resources including more than 10,000 plugins built by users, as well as hundreds of local chapters and more than 650 community-organized events in a single year. Notion hired a community manager early on and built a community program more than 300 ambassadors around the worldto the point where a Notion event is happening almost every day somewhere in the world. Salesforce is running over 1,300 community group meetings in over 90 countries with a formal MVP recognition scale.
None of this is a side project. Everyone has resources, is owned and has room to maneuver. Compare this to the industry benchmark: In CMX’s 2025 Community Professional Survey, 30% were solo teams and 17% had no full-time community staff.the highest ever, while 37% cited “difficulty proving ROI” as their top budget challenge. (This survey comes from a community software company, so weight it accordingly, but the underfunding it describes is what you see everywhere.) The difference between thriving communities and abandoned communities is often less in the idea than in the investment behind it.
Set your expectations with the 90-9-1 rule
Before you build anything, do some research so you don’t panic at the first sign of calm. In 2006, Jakob Nielsen documented inequality of participation: In most online communities, about 90% of people hide, 9% contribute occasionally, and 1% produce the bulk of the activity.
This is a heuristic, not a law, and the exact ratios vary widely. But it reframes what success looks like. A community of 1,000 people where 10 people post daily is not a failure; This is the model that works as expected. Founders who expect everyone to participate are impatiently killing healthy communities. Design your community with lurkers in mind, as they continue to read, build trust, and often become customers.
Choose a job and instrument it
The most common failure after understaffing is trying to force a community to do everything at once. David Spinks, who founded CMX and wrote the book on community strategy, offers a framework worth stealing: A community can drive multiple distinct business outcomes, support, product feedback, acquisition, retention, etc.and his direct advice to small teams is to focus on just one or two, unless you want to fail.
This focus is also what makes ROI provable, which solves the problem that killed the budget in the first place. If your community’s job is to support, measure ticket deflection. If it’s retention, measure the difference in churn rates between members and non-members. If it’s a product, count the features delivered from community feedback. CMX data revealed that teams with community data connected to their CRM were about twice as likely to rate their community as “extremely successful”because they could finally see the line between engagement and income. Choose the outcome before you build the space and measure it from day one.
What to do this quarter
You don’t need 300 ambassadors to get started. You need the first ten real members, why they would show up, and a metric that ties their presence to your business.
Start absurdly small and personal. Founders should actively participate in the community, especially early on, helping establish the culture before it can thrive on its own. Give people a reason to come back that isn’t your product: a place to get better at the work your product supports, answers they can’t easily find elsewhere, peers on their level. Recognize your 1%, because the handful of people who do most of the activity are your most valuable asset and they continue to be seen. And resist the urge to broadcast; a community where the company speaks more than the members is a mailing list with additional steps.
An honest caveat: community-led growth is slow and it’s not the right first channel if you need a pipeline next month. It accumulates over years, not weeks, which is exactly why so many startups abandon it right before it has started working.
2026’s reason to bother
There is a timely argument for doing so now. As AI floods every channel with competent, forgettable contentand CMX found 81% of community teams already use AI toolsthe rare and unautomatable thing is a group of real humans who know and trust each other. A model can generate the form of expertise. This cannot generate the fact that a community member helped you last Tuesday and will do so again next week.
This is the real competitive advantage. Built like a product, owned properly, focused on one job, and measured honestly, a community is one of the few growth engines that becomes more defensible as everything else becomes easier to fake. Built after the fact, it is just one cemetery among others awaiting its logo.
Image by pch.vector on Magnific
