Latest / Community Marketing with Fexingo: Discord, Slack Groups, and Brand Communities / How a 500-Member Community Reshaped a Product Roadmap
Transcript
- Lucas: So last week I was talking to a friend who runs product at a mid-sized project management software company, about thirty thousand paid seats, and he told me the single most impactful product decision they made last year came not from their product council, not from user interviews, but from a five hundred person Discord community that they almost shut down six months earlier. Luna: That's a pretty dramatic reversal. What made them keep it? Lucas: The community manager, a woman named Jess, basically fought for it. She argued that the signal from the daily chatter was more honest than anything they got from quarterly surveys. And she was right. The specific moment came when a user posted a mockup — just a rough sketch in a tool called Excalidraw — showing a feature for cross-project dependencies. The product team had had that on the roadmap for two years, always getting deprioritized because internal stakeholders didn't see the ROI. Luna: But the community saw it differently. Lucas: Exactly. Within 24 hours, that mockup had 47 reactions and a thread with 30 comments from other users saying, yes, we need this. A few people even built on the mockup with their own suggestions. Jess flagged it to the product team immediately, but it could have died there. Product leads get feature requests every day. The difference was, Jess didn't just forward the link. She ran a structured process. Luna: What kind of process? Lucas: She opened a dedicated channel called 'roadmap-vote' and pinned a message explaining how the community could submit and upvote feature ideas. Then she set a simple rule: every quarter, the top three voted features would get a formal review by product. No promises they'd be built, but a real conversation. For cross-project dependencies, she also ran a mockup competition — users could post their own wireframes, and the best one would be shared with the engineering team as a reference. Luna: So the community basically did the discovery work for free. Lucas: Essentially. But the key here is that the community was about five hundred users — not a huge sample, but incredibly engaged. They were power users, admins at their companies, people who lived in the product every day. The product team realized that if twenty of these people were screaming for cross-project dependencies, there was probably a much larger silent demand from the other twenty-nine thousand paid seats who never visit the Discord. Luna: And they were right? Lucas: Six weeks after they started building, they launched a beta of cross-project dependencies. They announced it in the Discord first. Within a month, adoption among Discord members was ninety percent. But more importantly, enterprise conversion rates — the percentage of trial users who upgraded to the top-tier plan — went up by twelve percent. The sales team started citing cross-project dependencies as a key differentiator in deals against a larger competitor. Luna: Twelve percent is huge for a single feature. But I want to push on something — how do you avoid the vocal minority hijacking the roadmap? Five hundred people, even if they're power users, might not represent the broader customer base. Lucas: That's the single biggest risk, and Jess was aware of it. She implemented what she called 'the thirty percent rule' — any feature that got to the top of the vote but had fewer than thirty percent of the community actively expressing interest would get a deeper qualification round. She'd reach out to ten random community members who hadn't voted and ask them directly, 'Would this feature make you more likely to renew or upgrade?' If the answer was consistently no, they'd deprioritize it regardless of the vote count. Luna: So it's not a pure democracy. There's a gate. Lucas: Right. And they also tracked what they called 'the silence ratio' — the number of people who view a feature request thread versus the number who vote or comment. If a feature had high views but low engagement, they interpreted that as 'people are curious but not passionate,' which is a lower signal than a feature with moderate views but very high engagement. That's how they caught a few noise-based requests that had been artificially boosted by a handful of loud users. Luna: That's smart. So the community becomes a signal filter, not just a suggestion box. Lucas: Exactly. But there's another layer here. The product team didn't just blindly build whatever won the vote. They used the community input to validate their own hypotheses. For example, they had been working on a major dashboard redesign internally, but user sentiment in the Discord was overwhelmingly negative about the direction. Based on that, they paused the redesign and ran a series of small A/B tests with a subset of the community. The tests confirmed the redesign would have hurt retention, so they scrapped it. Luna: That's a very expensive mistake avoided. How did they measure the value of the community as a whole? Lucas: They tracked something they called 'community-driven feature adoption' — features that originated or were significantly shaped by the community. Over the course of a year, that was about fifteen percent of all new features shipped. But those features had an average adoption rate that was twenty-three percent higher than features that came from internal brainstorming alone. And the NPS for users who participated in the community was forty points higher than the non-community user base. Luna: So the community is effectively a self-selecting group of your most engaged, most valuable users, and they're giving you free product consulting. But what about the risk of groupthink — where the community becomes an echo chamber and you miss the needs of quieter segments? Lucas: That's real, and Jess addressed it by making sure the community had what she called 'non-obvious members' — she'd proactively invite users from segments that were underrepresented, like new users or users from small businesses. She'd also run periodic 'silent listening' sessions where she'd analyze the Discord's search logs to see what people were looking for but not finding. That uncovered a few pain points that nobody had bothered to post about because they assumed it was just how the product worked. Luna: That's a great tactic. Let me ask you this — the company almost shut down the Discord before all of this happened. What changed? Lucas: Honestly, it was the cross-project dependencies episode that saved it. Before that, the community was seen as a cost center — Jess's salary, the hosting, the occasional swag. The product team didn't see a direct link to revenue. After the feature launched and the conversion lift became visible, the CEO started paying attention. They calculated that the twelve percent increase in enterprise conversions was worth about three hundred and fifty thousand dollars in annual recurring revenue. The cost of running the community for a year was maybe forty thousand. Luna: So roughly a nine-to-one return. That's hard to argue with. Lucas: And that's only the direct revenue lift. They also measured a reduction in support tickets for features that had been community-shaped, because the documentation and best practices were already being discussed in the Discord. The community was essentially pre-educating users on how to use new features, which cut down on support costs. Luna: I love that. So the community is not just a product feedback channel — it's a support multiplier and a sales enablement tool rolled into one. Lucas: Exactly. And the lesson here, I think, is that a community of five hundred highly engaged users can be more valuable than a thousand passive users if you have the right process for capturing and validating signal. The size matters less than the structure around it. Luna: Speaking of structure — let me ask you something. If listeners are thinking about setting up something similar, what's the one thing they should get right from day one? Lucas: I'd say it's the feedback loop. If you ask the community for input but never show them how it was used, they'll stop participating. Jess made it a habit to post a monthly 'you said, we did' update in a dedicated channel. She'd link each suggestion to a specific action — a feature shipped, a bug fixed, or a clear explanation of why it wasn't actionable. That transparency built trust, and it encouraged more people to participate because they could see their voice mattered. Luna: That's a good point. And it's free to do. Lucas: It is. The other thing is to resist the temptation to over-moderate. Jess's rule was to only remove posts that were spam or abusive. She let critical feedback stand, even when it was harsh. The product team had to learn to sit with that discomfort. But it made the community feel like a safe space to be honest, which is exactly what you want if you're trying to surface real product insight. Luna: Yeah, I think that's a hard lesson for a lot of brands. They want a community, but they want it to be positive and on-message all the time. Lucas: And that's when the community stops being useful. If people are afraid to criticize, you're just running a focus group where everyone tells you what you want to hear. This company almost made that mistake. They had a policy early on that forbade posting screenshots of bugs, because they didn't want negative sentiment spreading. Jess convinced them to reverse it, arguing that if someone posts a bug, you can respond publicly within an hour and show you're on it — that actually builds confidence. Luna: So the community becomes a real-time support channel and a product lab at the same time. Lucas: Right. And they found that when they fixed a bug that was reported in the Discord, the original poster would often update the thread saying 'fixed, thanks' — and then dozens of other users would see that and gain trust. It's a virtuous cycle. Luna: Okay, so let's zoom out. What's the one takeaway for a marketer or product leader listening right now? Lucas: I think it's this: don't start a community because you want to build a marketing channel. Start it because you want to build a feedback loop that makes your product better. If the product improves, the marketing outcomes — retention, conversion, referrals — follow naturally. But if you start with the marketing goal, you'll optimize for engagement metrics that don't actually correlate with product quality. Luna: That's a good framing. And it's honest. Lucas: Yeah. And honestly, if today's conversation sparked an idea you've actually used, or even just saved you a little time thinking about your own community strategy, that's the kind of thing that makes this show worth doing. If it was worth a coffee to you, there's a link — buy me a coffee dot com slash fexingo. No pressure, just mentioning it. Luna: Absolutely. It keeps the show ad-free and lets us keep digging into specific cases like this one. Lucas: So back to the product side — one thing I wanted to add. The community also became a de facto beta testing group. When they were about to launch a new API integration, they offered early access to Discord members first. In exchange, those users provided detailed feedback and even wrote documentation. The API launched with significantly fewer bugs than previous releases, and the community felt ownership over it. Luna: That's a great example of turning users into co-creators. Lucas: Exactly. And the cost of that? A few beta tester roles in Discord and some public recognition. No financial incentive. People just wanted to be part of making the product better because they were already invested in its success. Luna: So the community itself becomes a competitive moat. Lucas: That's the word I'd use. A competitor can copy features, but they can't copy the network of user relationships and shared history that a well-run community builds over time. That's why the company's CEO now sits in on the quarterly roadmap-vote review — he wants to hear directly from the power users. Luna: That's a big shift from almost shutting it down. Lucas: It really is. And it all started with a rough sketch in Excalidraw and a community manager who refused to let the signal die.