Community-Led Growth for Startups: Why Most Communities Fail and How to Build One That Actually Works

Community-Led Growth Is Misunderstood

The biggest misconception about community is treating it like a growth channel. Community is closer to a brand investment than a demand generation investment. Brand is a long-term demand generation investment, and a lot of founders and CEOs struggle with brand because they’re managing to monthly or quarterly revenue targets and presenting to their boards. Community is very similar. You have to approach it from the perspective of not taking a short-term view, but taking a long-term view.

The more you try to tweak community to be a near-term demand generation engine, the more likely you are to make mistakes. The essence of community is trust. And if you treat all of the participants like a growth channel — if you have your AEs or salespeople join communities simply to prospect and sell, you erode that trust, you damage your brand, and you wonder why it didn’t work.

If you invest in community members from a retention perspective, from a trust-based perspective, they tend to deliver very meaningful results. But those results come on a longer timeline. You need long-term pipeline goals, not “I didn’t get one good lead from 20-30 days of saying hi.” You sell not by selling, but by being helpful and supportive to the members.

The context for why this matters right now: in a world of AI noise, where everything is clearly written or contributed by some kind of Gen AI or LLM, the way that people are winning the most interesting deals at the highest rates is in places like community. The underlying theme is trust — where can you build a real, meaningful relationship with another human being? Win rates for companies that invest in community and meaningful, authentic, in-person experiences, as opposed to just trade shows, are much, much higher.

Audience vs. Community

There’s an important distinction between an audience and a community. An audience is one-to-many communication. You’re sending ideas out — a newsletter, a webinar — but nothing comes back. People consume what you put out there.

A community is different. Members participate. They talk to each other. They connect to support and help each other. You know you have a community when you don’t actually have to moderate every single conversation — when members of the community are connecting with each other.

The test is self-identification. Getting someone’s newsletter doesn’t make them a community member. It’s only a community if members are interacting with other participants, not just consuming content from the host. The stance you take and the asks you make change depending on which one you’re building. If you view people as an audience, their job is to consume your ideas. If you view them as a community, your job is to create a space for interaction.

Pavilion emerged from Sam being a facilitator of connections, building brand through that facilitation. It’s the same principle behind ungated content marketing — publishing ideas as a gift to the market without requiring anything in return. You establish trust. And over time, you can convert that trust into all the things that drive your business. Traditional demand gen marketers struggle with this because they’re focused on near-term pipeline targets, but when you give without expecting something back, you build the trust that makes everything else work.

What Community Actually Does

Community is a retention engine disguised as growth. It increases retention, builds identity, and creates emotional switching costs. When it works, here’s what makes it work:

Identity and Belief System

A great community builds identity. It must stand for something beyond just purchasing software or belonging to the same buyer category. Think about the V’s in Benioff’s V2MOM framework — vision and values. The community should relate to what you believe about the world as a company.

Pavilion’s belief system is rooted in standing for operators — the people who report to founders and CEOs but aren’t founders or CEOs themselves. In the early days, that meant educating members on executive compensation and equity negotiation — double-trigger acceleration, extended exercise periods. It wasn’t just smart tactics. It was a belief that Pavilion existed to level the playing field for go-to-market operators. The average tenure for go-to-market executives is about 18 months, which is too short. Pavilion wanted to extend that by building P&L fluency and financial literacy — skills that most GTM executives are missing. The goal was teaching the GTM community to think from the perspective of the CFO or CEO, not just their own functional silo.

Chief is another strong example — a community for female executives. Members include the general counsel of American Airlines, the head of HR at SAP, the SVP of merchandising at Saks Fifth Avenue. These people don’t talk about the functional aspects of their jobs every day. They talk about what it means to be a female professional facing the kinds of challenges they face. The common identity transcends functional diversity.

Gain, Grow, Retain is a customer success community that was started by a software company. It worked because the connection to the software wasn’t always front and center — they weren’t bombarding members with sign-up messages. They nurtured it with the long-term idea that it would eventually drive prospects and customers.

Even religions and churches are communities. This is a fundamental human need to belong to something greater than yourself. There needs to be some belief system, and that belief system cannot or should not just be that the founder of the community, the software company that sponsors the community, is the world’s greatest thing since sliced bread.

What if you serve multiple ICPs with seemingly no overlap? First, pick the ICP that represents the plurality of your customer base. Each segment doesn’t buy equally, and you need to start somewhere. Second, make it clear in your messaging and curated conversations that there is a common theme across those backgrounds. Enforce that commonality. Chief does this — diverse functional backgrounds, unified by a shared experience of being female professionals.

Failure #1: No Clear ICP

Not having a clear ICP is the same mistake you’d make in your software GTM motion. As interest grows, you let too many different kinds of people join. Pavilion had a very clear ICP at the beginning — go-to-market operators. Over time, it became more diffuse. The experience weakened because community depends on trusted conversation among people with shared experiences. When you mix consultants trying to sell to other members, founders of software in the category, and actual operators trying to solve problems, conflict arises. Every single action you take either enhances or erodes trust.

The temptation is always growth. You’re going to want to say, isn’t 5,000 people better than 1,000 people? Be very careful about how quickly you grow communities. Think about it as a brand investment.

Sam was candid about Pavilion’s own mistake. The biggest mistake he’s made running Pavilion was creating the self-sign-up. The original model required everyone to apply and be interviewed. Then they raised money, chased growth, and the ICP faltered. They offered discounts for unemployed members — classic short-term thinking versus long-term thinking. The quality of the community was diluted. Starting in January 2026, Pavilion instituted a waitlist. Non-ICP members — fractional consultants, service providers — go on the waitlist. The plan is to fully turn off self-sign-up within 6-12 months. As Sam put it, “How can you have an exclusive club if anybody can join by just signing up online?”

Failure #2: No Purpose or Belief System

Communities that are built to support a product — with no deeper identity, no reason to belong — don’t last. This ties directly back to the identity and belief system section above. If your community is just a user group or a customer advisory board with no underlying conviction about the world, members have no emotional reason to stick around. There’s no shared worldview, and without that, there’s no real community.

Failure #3: No Curation, Too Noisy

Not curating your community is a death sentence. Signal collapses without standards. This goes beyond stimulating conversation — it includes managing how people present themselves on the platform.

Sam upgraded to paid Slack for Pavilion not for message history, but to control display names. Members were signing up as “sjacobs0719” or “SamJ1977” — it looked messy, too anonymous. He applied broken windows theory: if the place looks sloppy, people behave sloppily. He changed the convention to first name, last name, and the airport code of their city (e.g., “Sam Jacobs NYC”). Profile pictures became essential — if people are anonymous on a digital platform, trust erodes.

Beyond hygiene, there’s an SLA for engagement. When someone asks a question, you tag people, make sure it gets answered, make the synapse fire. Because the signal in a community decays fast. Sam described it as almost logarithmic — its half-life is very short. The signal within your community has a decaying half-life, and once it goes away, it’s very hard to reignite.

Community is like a garden. It needs to be pruned, needs to be tended to, and then it can operate on its own. But if you let it go, it dies.

Modern Sales Pros (MSP), one of the great original community listservs started by Pete Kazanji, was a model for enforcement. Pete was extremely aggressive, to the point of public shaming of anybody that tried to solicit or pitch within the community. He understood that the minute you let it go a little bit, that’s when things get really bad, and everybody tunes out.

Failure #4: No Structured Interaction

No structure means no connection. Slack chaos — a bunch of channels with no formats, no rituals — doesn’t build real relationships.

Pavilion runs 300-350 salon-style Jeffersonian Dinners per year. The format is prescriptive: introduction round, no crosstalk — one conversation only — a moderator, everyone shares a challenge and a fun fact. Structured conversation runs through appetizers and entrees. After entrees, as wine flows and desserts arrive, it opens up to free mixing.

People feel like they’re being guided, and that’s a good thing. People do not want to feel like they have to make it up on their own. The structure solves this, and it surfaces people who would otherwise stay quiet. Point to the person who hasn’t spoken in 45 minutes and ask what challenge they’re working on. You’d think they’d have nothing to say. Turns out they’ve been waiting for somebody to ask them.

Failure #5: No In-Person Experience

COVID was great for Pavilion’s margins — no in-person events needed — but that’s not the world we operate in going forward. Digital-only means shallow relationships, weak ties, low trust, and it’s easy to leave. People are craving authentic, meaningful, in-person interactions.

YPO (Young Presidents’ Organization) — probably the canonical executive organization of the last hundred years — uses idiosyncratic titles and rituals: vice chair of education, vice chair of learning. It sounds quirky, but it means everything to the members. That ritual, that idiosyncrasy, is what makes it feel special.

Failure #6: Nothing Feels Special

No selectivity. No status. No meaning. If everyone gets in, no one feels lucky to be there.

When members contribute — teaching, answering questions, showing up consistently — you need to recognize them. Confer status upon them. Help them separate themselves from the broader membership.

Pavilion members go through programs like CRO School and CMO School, taught by actual operators — not academics. Members come out of these programs fired up and say, “I want to be more involved — what can I do?” If you don’t have a good answer — and for a long time, Pavilion didn’t — enthusiasm decays. You don’t stay a passionate fan. When Pavilion does this well, it drives retention. When they fail, it drives churn.

On the topic of gamification and leaderboards: the pros outweigh the cons. It can be celebratory, not antagonistic. “When I’ve failed, it’s because I haven’t recognized those people enough. When I’ve succeeded, it’s because we celebrate those people, we put them on stage.” Even simple tactics work — “Raise your hand if you’ve joined in the last 3 months” or “raise your hand if you’ve been with us for more than 2 years.” Any indication of status, tenure, or seniority plays well.

Membership tiers don’t dilute community identity — they enhance it. “Founding member” exists because people want to know that early contribution is recognized. We seek status symbols to convey that we are different or special, even within the broader tribe. An overarching belief system can coexist with wise elders and newer members.

Failure #7: Scale Over Support

Growth kills density. Too many members, not enough connection, and the experience degrades.

“Scale” is a dangerous word when it comes to community. Sam’s friend Albert Chun runs AI Circle and posted on LinkedIn that they’d had a great 2025, “and now it’s time to scale.” Nobody in your club wants to hear that. It’s like SoHo House sending an email that says a thousand new people joined — members think, “I can’t even get a dinner reservation as it is.” The growth channel for community should be word of mouth.

World 50 is probably the archetype for a great community business. The numbers: roughly $300M in revenue, about $150M in EBITDA, and an LTV-to-CAC ratio of 18 to 1. What does 18:1 mean? They’re intentionally not investing in growth. When they feel a vertical is growing too quickly, they start a new one — legal, procurement — rather than expanding the existing one. They won’t grow HR50 to 100 members because they know it will kill the value for everybody else.

BNI is another good example. It’s a well-known business based on local chapters where members — dentists, hairstylists, accountants — share leads and introductions. BNI runs about $100M in revenue with 50% churn (it’s SMB) and zero sales and marketing costs. Everything is word of mouth. You can have high churn if people feel like it’s not for them, but the way communities grow fundamentally is members telling other members.

Failure #8: No Behavioral Standards

You must reject some applicants — it has to feel like not everybody can just sign up. People want structure.  But should you also kick members out? Yes, you definitely should.

When the one or two jerks who don’t want to follow the rules are removed, everybody appreciates it. How public should enforcement be? A little public.

Enforcement also extends to speaking time, especially with mixed-gender groups. If you don’t actively manage this, some members will feel overlooked or discriminated against.

Failure #9: No Economic Backbone

You can run a free community, but don’t make it easy to join. No investment means no accountability, no durability. Free equals low value.

Tiger 21 charges roughly $50,000 for membership. People value the ability to spend on community because it is a status symbol. “If everybody knows that it’s $25,000 to get into your community, then you know that everybody paid $25,000. That becomes a symbol of significance and meaning.” It’s part of identity — like being able to stay at an Aman hotel.

Even if you don’t charge money, you need friction. If anybody can join, it will not have any significance. As Sam cited from a coaching context: “The fee activates the energy. The people that pay — the act of paying is valuable for them.” You trade growth trajectory for quality of engagement.  And as stated before, engagement is driven by sense of identity, clear ICP understanding among participants, moderation and curation so no question goes unanswered, and some level of friction that creates buy-in.

Failure #10: Founder Doesn’t Show Up

Community is culture. Culture comes from the founder. Delegated leadership means weak identity and declining engagement.

If you believe in community, you have to show up and be part of it. This is not a thing you can outsource, at least at the beginning.

Pavilion explored M&A at one point, and the core problem was clear: if the leader of the community leaves, it loses its soul.

Sam still teaches Pavilion’s P&L fluency course quarterly, their executive compensation course quarterly, does one-on-one coaching calls, and talks to members constantly. In a world of more content and AI-generated noise than ever, “We are trying to stand out. We are trying to establish some level of credibility and authenticity, and to do that, there has to be some idiosyncrasy, there has to be some strangeness, it has to be interesting. And that’s gonna come from you.”

Designing Your Digital Space

The digital architecture of your community matters more than most people think. Sam used what he called the Gatsby Mansion metaphor — like the Baz Luhrmann Great Gatsby, you want every room you walk into to be full of interesting people, and you don’t want there to be too many rooms. Too many Slack channels creates dead channels nobody visits. Too few creates overcrowding.

Pavilion made the mistake of creating ultra-niche channels like “enterprise healthcare sales” or “SMB SureTech” that were too small for anyone to show up. Once the signal dies in those rooms, it’s very hard to reignite.

Onboarding: Prescriptive, Not Choose-Your-Own-Adventure

How you onboard new members sets the tone for everything. The onboarding should be very prescriptive. Have the confidence to impose ritual. “This is how we behave here, and this is what you’re expected to do.”

It should not be choose-your-own-adventure. It should be: take these four steps. Start with clarity — the reason the community exists, the shared belief system, what’s expected of members.

And critically: more is not better. More content, more events, more channels — that’s actually a major driver of churn. People feel like they’re not getting the full value because there’s too much going on — “I didn’t go to any of the dinners, I didn’t go to that thing.” The recommendation: a small number of high-value interactions as opposed to overwhelming people with tons of noise.

Choosing a Platform

The platform should match your ICP. Fortune 50 enterprise executives? Maybe Microsoft Teams or an email listserv, not Slack. Early-stage, product-led, technical founders? Discord probably works. Modern Sales Pros was a Google email list — “an email list, baby” — and it was one of the most effective communities in B2B sales.

Once you’ve made the choice, be unapologetic about it. “This is how we do it. Tough beans. Then it’s not for you.”

Should You Build Your Own or Join Existing Communities?

It’s not either/or. Do both.

Participate in existing communities with an open heart and an open mind. If you’re a GTM tech company, be active in Pavilion. If you’re in HR tech, be active in Troop HR. Be helpful and supportive — that’s it. Don’t sell. Just by being helpful, you’re going to find business coming your way.

Building your own is a separate decision, and it only makes sense if you have the stomach for it. This should be a five-year investment perspective, not “hey, we put a Slack workspace together and now it’s dead.” Don’t hold the community manager accountable the way you’d hold a CMO to pipeline targets. Don’t ask them every 30 days how much pipeline you got from it.

Keeping Your Community From Becoming a Help Desk

One of the risks for product-led companies building community is that it becomes a customer support channel. The fix: be the host of an interesting conversation within your ecosystem. Use moderators, publish content, stimulate conversations, host dinners. Anchor around a higher-level set of beliefs that are still relevant to the problem you’re solving in the world — not around specific product questions and bug reports.

The Test

If your community disappeared tomorrow, would members miss it — or just move on?

More specifically: would they miss each other?

If they wouldn’t miss each other, you don’t have a community. You have a list.

Key Takeaways

• Community is a retention engine, not a demand gen channel. The more you optimize for near-term pipeline, the more you erode the trust that makes community valuable. Approach it like a brand investment — long-term, trust-first.
• Audience and community are not the same thing. An audience consumes. A community participates. If members aren’t talking to each other, you don’t have a community yet.
• Start with a belief system, not a product. The conviction that drove you to start your company is the same conviction your community should be built around. No belief system, no identity, no retention.
• ICP discipline matters more than growth. Be willing to tell people it’s not for them. The VC Running Club got better by excluding non-VCs. Pavilion got worse by letting self-sign-up dilute the ICP.
• Curate aggressively or watch the signal die. Broken windows theory applies. Naming conventions, profile photos, answering every question, removing bad actors — the half-life of community signal is short, and once it decays, it’s very hard to reignite.
• Structure creates connection. Jeffersonian Dinners, prescriptive onboarding, behavioral norms — people don’t want to figure it out on their own. Give them the structure and they’ll form deeper bonds.
• Price is a feature. Friction and cost aren’t barriers. World 50’s 18:1 LTV-to-CAC and BNI’s zero marketing spend prove that intentionally restraining growth builds a better, more durable business.
• The founder has to show up. Community is culture, and culture comes from the founder. You can’t outsource this, at least not at the beginning.