Latest / Community Marketing with Fexingo: Discord, Slack Groups, and Brand Communities / When a 200-Member Community Sourced 50 Product Features in 30 Days
Transcript
- Lucas: Flowtide is a productivity app — think task management with a social layer. In April this year, they had a 200-member private community on Circle, and they decided to run a 30-day feature sourcing challenge. The goal: generate as many viable product ideas as possible from the community itself. Luna: And the result? Fifty ideas, fully documented, in thirty days. That's a rate of almost two per day from a group smaller than most high school classrooms. Lucas: Exactly. But what's more interesting is how they did it. Because just saying 'send us your ideas' would have gotten maybe ten submissions, mostly noise. Flowtide used three specific tactics that turned a small group into a feature factory. Luna: Let's start with the first one. I read about this in their community manager's write-up — daily prompts, but not the usual 'what do you want' kind. Lucas: Right. Every day at 9 AM Eastern, they posted a prompt tied to a specific workflow. 'How do you currently handle recurring tasks?' or 'What's the one thing that makes you abandon a project mid-week?' These weren't feature requests — they were problem statements. The community answered in threads, and the product team mined those threads for underlying patterns. Luna: That's smart because it removes the bias of users proposing solutions. Users tend to suggest features they've seen in other apps, not necessarily what would solve their actual pain points. Lucas: Precisely. And the second tactic: they used a voting mechanism that didn't allow ranking. Members could upvote as many ideas as they wanted, but they couldn't give a '10 out of 10' or a star rating. The team measured raw signal — how many people cared enough to click, not the intensity of a few loud voices. Luna: That avoids the vocal minority problem. In a 200-person group, you might have fifteen people who dominate every thread. But with binary voting across the whole group, you get a truer picture of what the majority actually needs. Lucas: And the third tactic was transparency. Every Monday, the product manager posted a public roadmap update showing which of the previous week's ideas had been moved into 'under consideration' or 'planned'. Not just a 'thanks for your feedback' — a specific reason why an idea was shelved or fast-tracked. Luna: I think that's the piece most companies miss. If you ask for input and then disappear into a black box, people stop contributing. Flowtide's PM responded to every single submission within 24 hours. Not with a canned reply — with a personal note. That kept engagement at sixty-eight percent over the full thirty days. Lucas: Sixty-eight percent is remarkable. Most brand communities see engagement drop to single digits after the first week of any campaign. The fact that two-thirds of members were still actively submitting and voting in week four tells you how much that personal response mattered. Luna: But there's a flip side. Fifty ideas is a lot. How many did they actually build? Lucas: Good question. As of last month, they've shipped twelve of them. That's a 24 percent conversion rate from idea to production. But here's the nuance — some of those ideas were small, like adding a keyboard shortcut. Others were substantial, like a new recurring task logic that required weeks of development. Luna: Twenty-four percent is actually high for crowdsourced features. Most internal product backlogs have a conversion rate closer to ten percent, and that's with professional product managers steering the ship. Lucas: Right. And the ideas that didn't get built weren't wasted. The community saw their contributions acknowledged, even if they weren't acted on. That builds trust. One member wrote in the community, 'I'm disappointed my idea didn't make it, but I understand the reasoning. I'll keep contributing.' That's the kind of loyalty you can't buy with discounts. Luna: So what's the cost side of this? Moderation, for one. Someone had to read and respond to every submission within 24 hours. That's not free. Lucas: The community manager said she spent about two hours per day on read and response. That's roughly forty hours over the month — one work week. Add the product manager's time for the weekly roadmap update, maybe another ten hours. So fifty hours total. Compare that to a traditional focus group: a single two-hour session with a moderator costs around ten thousand dollars and yields maybe a dozen usable insights. The community approach cost roughly a week of salary for two people and produced fifty documented ideas. Luna: That's a compelling ROI. But there's another hidden risk: feature bloat. When you invite your community to suggest features, you're implicitly telling them you'll consider everything. If you don't manage expectations, you can end up with a bloated product or a disappointed community. Lucas: Flowtide handled that by being explicit from day one. The challenge description said, 'We will build the ideas that align with our product vision and have broad support.' They also capped the total number of ideas they would accept — not a hard cap on submissions, but a soft cap on what would be reviewed each week. That kept the scope manageable. Luna: And the silent majority? We've talked about them in previous episodes. How do you know you're not just hearing from the most engaged twenty percent? Lucas: That's the perennial question. Flowtide did a few things. First, they randomized the order of prompts so different members saw different problems on different days, reducing fatigue. Second, they sent a weekly email to every member — including lurkers — with a one-click survey: 'Rate how satisfied you are with the current feature set.' That gave them a pulse from people who never posted. The results aligned closely with the voting data, which gave them confidence that the community's voice wasn't just the loudest one. Luna: So the model works for a 200-person community. But what about scaling? If Flowtide grows to 2,000 members, does this still work? Lucas: That's the open question. Flowtide's community manager told me they're experimenting with a tiered system for the next challenge. They'll have a core group of 200 'power contributors' who get the daily prompts, and a larger group of 2,000 who get weekly summaries and quarterly surveys. The idea is to preserve the intimacy of the small group while expanding the feedback surface area. Luna: That echoes what we've seen in other successful communities — the 100 true fans model, but applied to feedback instead of revenue. You don't need everyone to be a super-contributor. You need a dedicated core and a mechanism to check that they're still representative. Lucas: Exactly. And speaking of what makes this model sustainable — it's worth noting that this kind of deep, ongoing conversation is exactly what listener support makes possible on our end. If these marketing deep dives have sparked something you've actually used in your own work, that's the whole point. And the way we keep this ad-free is listener support at buy me a coffee dot com slash fexingo. Luna: Yeah, it's a small gesture that keeps the podcast independent and lets us dig into cases like this without worrying about sponsors. Lucas: So back to Flowtide — the bigger lesson here might be about the shift from 'community as audience' to 'community as R&D department.' Most brands still treat their communities as a broadcast channel for updates and promotions. Flowtide flipped that script and turned their community into a product development engine. Luna: And they did it with 200 people. That's the part I keep coming back to. You don't need millions of users to get high-quality product input. You need the right structure and the right respect for people's time. Lucas: Respect is the right word. Every prompt, every response, every roadmap update communicated: 'Your time is valuable, and we're using it wisely.' That's rare. And I think that's why the community stayed engaged for thirty days. Luna: So if someone listening wants to try this, what's the one thing they should do differently from Flowtide? Lucas: I'd say start smaller. Don't launch a 30-day challenge out of the gate. Run a one-week pilot with your most active twenty members. See if the daily prompt rhythm works, see if you can keep up with responses. Then scale from there. The worst thing you can do is overpromise and under-deliver on community feedback. Luna: Good advice. And maybe make sure you have a product manager who actually enjoys talking to users. Not every PM does. Lucas: That's a whole other episode. But for now, the takeaway is this: a small, well-structured community can outperform a big budget focus group. The tools are cheap. The hard part is the discipline of listening and responding.