How to Create a Vibrant Community

Nobody assigns you a community. You don't get a budget, a mandate, or a room full of people who already like each other. You get a channel, half an hour on someone's calendar, and a suspicion that a lot of smart people are solving the same problems without knowing about each other.

Two years later, that suspicion was a community of 1,600+ people. It was never complicated. It was just relentless. This is the short version of what I did and why it worked — the handful of choices that actually matter, and the discipline of not giving up in month two, when the room is still small.

Why Bother?

The knowledge I wanted to connect was everywhere and nowhere at the same time. Every team was experimenting on its own, often unaware that people a floor away were working on the exact same problem. It was never a talent problem — there was plenty of talent. It was a communication problem, which is a much more boring thing to be, and a much easier one to fix.

The Whole Strategy, In One Sentence

“Connect the people already doing the work and get out of their way.”

That was the whole plan. Community-led instead of top-down. Open by default, so nobody had to ask permission to join and learn. It sounds too simple to be a strategy, and that simplicity was the point.

So What Do You Actually Do?

I kept the launch simple, partly on principle, partly because there was no budget for anything else. One 90-minute kickoff, recorded, because if it isn't recorded, it didn't happen. A shared hub page two weeks later. After that, I let the formats sort themselves out.

Two formats carry the community now, and each has exactly one job — a format that tries to do two usually does neither:

• Show & Tell — a few times a month, demo-first. People show the real work they've built. By a wide margin, the best way to bring new people in.
• The Brownbag — hands-on and informal. People come to build and try things out after the session, not just to watch.

None of this needed a big budget, a mandate, or a Chief Community Officer (not a title I have or want). It needed a calendar invite, a recording button, and enough perseverance to keep going in month two, when twelve people showed up instead of eighty, and quitting would have been the easy thing to do.

A Simple Method To Start

• Step 0 — Make a short list of people already enthusiastic about the topic.
• Step 1 — Reach out and find people already doing the work.
• Step 2 — Give them one clear reason to show up.
• Step 3 — Record it and make it easy to find.
• Step 4 — Do it again next week.

What Actually Works

The Show & Tell works because it does a few things at once. It makes ongoing work visible across teams. It gives people something concrete to react to. The sessions that land hardest are never the polished ones — they're the ones where one team shows what broke when they tried something another team had already tried, or walks through a failure that saved someone else three weeks. Practical beats theoretical every time.

The Brownbag works differently. Watching a demo isn't the same as doing the work, so the Brownbag is hands-on — people leave able to apply something that same afternoon. One format brings people in; the other helps them get better once they're here. Run both, on purpose.

What If You Screw Up? (You Will.)

I did, more than once, in the open. A single monthly call became two, once it was clear that a broad Show & Tell and a core-team meeting where things actually get decided are not the same meeting. Announcements moved from one channel to several, because people read in different places and picking just one means half of them never hear you.

None of it was fatal. Fixing mistakes in public builds more trust than hiding them, because people believe a community that admits it's still working things out. When the organizers show the messy, half-finished work first, everyone else feels safe enough to do the same — which is the whole point.

The Numbers, Because Feelings Aren't Data

The totals matter less than the shape. Over the last nine months, attendance stayed in its range with no real drop-off, and membership kept climbing, month over month. That's the opposite of the usual pattern, where a community spikes at launch and then goes quiet. None of it happened because the topic was exciting — plenty of exciting things have empty channels. It happened because every format had one job, every session was recorded, and every question got a real answer instead of “let me get back to you.”

Is It Still Alive, Or Did It Just Spike?

Most of the nice things people say come straight to you, which is lovely but isn't really evidence. Two moments were. Almost two years in, a member shared one of our sessions in a channel none of us ran, unprompted, and simply called it worth watching. Separately, another member asked to fold the community into a bigger internal initiative so more people would see what was happening inside it. Nobody asked either of them to do it. Both happened long after launch, which is the only time that kind of unprompted pull tells you anything.

Does It Actually Help? (I Asked.)

Attendance tells you people keep showing up. It doesn't tell you whether showing up changes how they work. So, two years in, I ran a short impact survey and asked people directly. Nine answered. That's a small, self-selected group, not a proper sample, and I'll treat it that way. But nine honest answers still tell you things a dashboard can't.

What people say they get lines up with everything above: a read on what other teams are doing, ideas they can take back to their own work, and a sense of which tools and approaches are actually worth their time. One person described the real value as reassurance — knowing whether it's just them, or whether everyone is wrestling with the same thing.

It also gave the first real answer to something I'd admitted I couldn't measure: I couldn't put a number on what the silos were costing. This is the closest I've come. One person first heard about a useful tool in a session, took it back to their team, and has been passing it on since. Another summed up the whole point in a line: “reduce duplication and increase adoption of best practices.” One good idea carried across a boundary that used to stop it is exactly the kind of waste a community is meant to remove. It's not a dollar figure, but it's real.

The survey didn't only flatter it, which is part of why it's worth trusting. When asked to boil the value down to a single number, two people refused outright — knowledge- sharing doesn't reduce to one KPI. One of the lower scorers said plainly they mostly see experienced people rather than newcomers. That's a fair criticism of the onboarding story, and I come back to it below. A survey you only quote when it agrees with you isn't worth much.

What This Is NOT Good At

Every design has trade-offs. I think these are worth it, and you should know them going in.

• It leans on individuals. A community that runs on a few committed people is fragile the week those people are on holiday. Rotating the core team helps, but it doesn't fully fix it.
• You can't prove the counterfactual. Silos are probably costing real duplicated effort, and the survey gave a first glimpse of it — but still not a clean number. Take “silos are expensive” as a well-supported hunch, not a measured fact.
• It rewards the already engaged. The people who show up are the people already inclined to show up. Mostly experienced people, not the newcomers you'd most like to reach. I haven't solved that one.
• It's slow to show ROI. For a long time it looks like a lot of calendar invites and not much else. You need patience, and a sponsor with patience, before the value compounds.

Where Will You Take It?

Your community won't look like mine, and it shouldn't. Different teams, different politics, different culture. The mechanics still carry over, though. Find the people already doing the work. Give them one clear reason to show up. Make the good material easy to find. Rotate the core team before anyone burns out. And be honest about what a community actually is: not a meeting series, but the fastest way I've found to make silos harder to sustain.