A community lead I worked with a while back had a slide she brought to every quarterly review. Member growth, posts, event attendance, satisfaction scores, all of it moving in the right direction, and she was rightly proud of it because the numbers were real and she had worked hard for them. Then somebody from finance asked what any of it had to do with renewals, and there was one of those pauses that lasts about four seconds and feels considerably longer, and then the meeting moved along to the next thing. She told me later that she hadn’t had a good answer, and that she’d been dreading that question for about a year.
I’ve been on both sides of that exchange. I’ve built that slide, more than once, and I’ve also sat in the room wishing the person presenting it would tell me something I could actually use.
The claim underneath the slide is one most of us in community work believe, which is that community helps with adoption, expansion, retention, and advocacy. I believe it too, and I’ve watched it happen at Asana, at Evernote, and across a lot of advisory work since. The problem is that “community improves retention,” said on its own, isn’t yet a claim about anything. It doesn’t say which customers, doing what, leading to what, over how long. It doesn’t hand anyone something to go and do differently next week. In a room where everyone else has a number attached to their name, it lands like a feeling rather than an argument, and people are polite about it and then move on.
So the question worth sitting with is the how. If community contributes to those outcomes once somebody has already become a customer, what is actually happening, to whom, and how would anyone know?
The question that helped at Asana
The community work at Asana was anchored in adoption, expansion, and retention for a long time before we could show any of it clearly. That’s what the ambassador program was built around, and the forum, and the events. By the end there were roughly 25,000 ambassadors, most of them team leads running Asana inside their own organizations, alongside a forum with hundreds of thousands of people in it and a global events program running into the hundreds of events a year. We knew what we were trying to influence, and we designed the programs with that in mind.
What we couldn’t do, for a long stretch, was see it. The question that eventually got us somewhere was a fairly plain one, and it came from the shape of the product. Asana is a team-based tool, so: are accounts with an ambassador in them adopting, expanding, and renewing differently from accounts without one?
Getting to an answer meant having community participation data sit next to account data, and that data was scattered across an array of disparate systems. Mostly what it took was going to our data analytics team, more than once, and asking them to help me connect those dots, with the particular mix of gratitude and persistence you develop when you need something from a team that doesn’t report to you. There wasn’t anything clever about the question itself. Answering it was the part that took time.
Once we could see it, the conversations with the rest of the business changed. We could compare accounts that had an ambassador in them against accounts that didn’t, look at how each group was behaving, and then talk about what to do with that. Usually it meant working with customer success to get ambassadors into the accounts where their presence would matter most, which is something a CS leader can plan around.
None of that is causation, and I’ve never claimed otherwise. It’s correlation. It was still worth having, because it told us where to put our effort, and I’d rather say that plainly than dress it up. Executives have heard the dressed-up version before, and once someone catches you overstating what you can prove, everything else you tell them gets discounted along with it.
Four different things wearing the same word
Part of what makes this hard is that “engagement” gets used as a single word for several processes that work quite differently from each other. Underneath it, the things community is actually doing after somebody becomes a customer tend to fall into four groups:
Adoption, where a customer actually starts using what they bought, helped along by seeing how a peer did the thing they’re trying to do.
Expansion, where a customer uses or buys more, either because a team went further with what it already had or because somebody carried a use case back to a different team.
Retention, where a customer stays, and part of the reason is that leaving would now mean giving up relationships, knowledge, and a bit of standing built alongside the product.
Advocacy, where a customer vouches for you, publicly or privately, because they’ve had enough value from being there to be willing to.
In practice they overlap constantly, and a community team, which is usually small, gets asked to influence all four at once. What differs is the mechanism underneath each one and the evidence that would show it working, which is why a single participation number can’t stand in for all of them.
The adoption one is the most familiar to anyone who has run a forum. A lot of the questions we got at Asana arrived looking tactical, about permissions or templates or how to structure a project, and underneath they were usually about how a group of people worked together. Somebody would set something up successfully on their own and then find that their team carried on doing it the old way, and no amount of documentation fixed that, because documentation explains how a feature behaves and not what happens when a dozen people have opinions about it. What tended to help was showing them how a team of a similar size and shape had actually made it stick.
Expansion covers a fair bit of ground. Sometimes it’s a team going further with what they already bought, picking up capabilities they’d ignored or moving from one workflow to several, which community helps with in much the same way it helps adoption. The version I find most interesting is the one that crosses into another team, partly because it doesn’t look like expansion while it’s happening. When I was at Gradual, one of our customers, an AI company, started their community program inside a single team with a deliberately narrow focus. Customers showed up and participated, and other teams inside that company could see it happening. Those teams wanted in, and over time they picked the platform up for their own purposes, first for customer education, then champion engagement, then customer marketing, and eventually demand gen events and webinars. Each team that came on board made the next one a bit easier, because there was more of it visible internally for somebody to point at. There were discovery calls and sales conversations along the way, the way there always are. What’s worth noticing is where the interest started, which was somebody inside that company seeing a colleague doing something adjacent and wanting a version of it for themselves. The commercial conversation followed that interest rather than manufacturing it, and that’s a meaningfully different thing from a rep working an account cold.
Where I’d stop short
Retention is where I’d be most careful, because it’s where the claims tend to get loosest.
The version I find plausible is about accumulation. A customer who has used the product for two years and done nothing else has one thing to lose if they leave. A customer who also knows a handful of people through the community, who has answered questions and been thanked for it, and who has picked up a way of working rather than a set of clicks, has rather more to rebuild somewhere else, at their own expense, on their own time. The software is replaceable in a way the rest of it isn’t, and the difference is closer to changing neighborhoods than changing banks.
That’s a reasonable story and it may well be true. It’s also very hard to separate from a much less flattering explanation, which is that the customers who accumulate all of that were the committed ones to begin with, and the community is where their commitment became visible rather than where it came from. I’ve never been able to fully pull those apart, and there’s one thing I’d want to know before believing the stronger version: is that depth showing up in accounts that didn’t already look like your best ones?
If it is, that’s the more interesting result, because something is happening in accounts where nothing was guaranteed, and it’s where I’d point the program next. If depth only ever appears in customers who were always going to stay, the program is still worth running for the people in it, and I’d stop describing it as a retention driver, because what you’re holding is a picture of your strongest accounts rather than an explanation of why they’re strong.
Advocacy is mostly a question of sequence. Think of it as a piggy bank: companies want to make a withdrawal before they’ve made any deposits. They want the referrals, the case studies, and the reference calls before they’ve put much into the relationship at all. Members can tell, which is why the ask so often lands with a thud and then somebody wonders why the advocacy program isn’t converting.
What happens when it works is less spontaneous than that might sound. Plenty of advocacy gets asked for, and there’s nothing wrong with that, because making asks is how advocacy programs actually run. What’s different is that by then somebody has gotten real value out of being there, some of it from the company and some of it from other members, and they’re willing to be asked.
The craft is in the asks themselves. They need to be the right size, made at a moment that makes sense for that person, and spaced out enough that being an advocate doesn’t turn into a second job. And whatever the ask is, an advocate can never be put in a position where they feel like they’re selling, because that’s the point at which they stop being credible to the people they’re talking to.
There’s a measurement trap here as well, and it’s easier to fall into than the retention one. Advocacy counts go up when you ask more often, so a program can look like it’s improving when what’s really happening is that the same willing people are being asked again. The number rises and the willingness drops, and nothing in the reporting shows you the second part until somebody stops replying.
So the version of the earlier question that applies here is this: how much of this year’s advocacy came from people you hadn’t already asked more than once? If most of it did, something is working. If it’s a small group being asked repeatedly, you’re spending down a handful of relationships, and the count won’t tell you when they run out.
Try this
Pick one of the four: adoption, expansion, retention, or advocacy. Write a single sentence naming the behavior you think matters for the one you picked, where you’d see that behavior happening, which outcome it plausibly influences, and who owns the next step. Then go and find out whether your systems can show you that behavior at all, because in a lot of companies that’s the actual blocker, and it’s better to discover it this week than in the middle of planning. If the measurement side is where you tend to get stuck, Why Nobody Cares About Your Community Engagement Metrics (and What to Do Instead) goes further into putting this in terms an executive will engage with.
Decoded Takeaways
Most people who work in or around community accept that it contributes to adoption, expansion, retention, and advocacy. That belief is reasonable and it’s also where the conversation usually stops, which is why the claim tends to fall apart the moment somebody outside community asks a second question about it. “Community improves retention” doesn’t say which customers, doing what, or leading to what, and it doesn’t give another team anything to act on.
The more useful version separates out the different things happening underneath the word engagement. Adoption runs on peer examples that do something documentation can’t. Expansion happens both when a team goes further with what it already has and when somebody carries a use case back to a different team. Retention plausibly comes from accumulated relationships and knowledge that raise what leaving actually costs. Advocacy works when an ask lands on somebody who has already received enough value to be willing to be asked. They overlap constantly, and most community teams are asked to influence all four at once, but each works through a different mechanism and shows up in different evidence.
The discipline that keeps all of it credible is being straight about what the numbers can’t show. At Asana, comparing accounts with an ambassador in them against accounts without gave us something real to work with, and it was never proof of cause. The same caution applies to advocacy, where the count goes up whenever you ask more often, so a program can look healthier at the same time as the willingness behind it is being spent down. Saying that out loud tends to make the rest of the argument easier for a skeptical executive to accept.
Related Posts
From our Partner
If you’ve been meaning to get hands-on with the AI side of GTM, GTM Engineer School’s fifth cohort starts 28 September.
It runs eight weeks, live. The first four are spent working in Clay, Octave, Sumble and Swan alongside the teams who built them. The back half is a capstone on whatever stack you actually run, finishing with demo days in front of hiring managers.
It’s $1,850, and every seat comes with $3,500 in platform credits for the tools. Enrollment closes 14 September.
Learn more at gtm-engineer-school.com.