Latest / Social Media Marketing with Fexingo: Instagram, TikTok, LinkedIn, and Platform Strategy / How Brands Use Social Media for Community-Led Product Development
Transcript
- Lucas: So there's this cosmetics brand — I'm going to keep the name out of it because the principle applies broadly — but they launched a new foundation shade last quarter that became their best-selling SKU within three weeks. Luna: Right, and I think I know the one you mean. They basically crowd-sourced the formula, didn't they? Lucas: Exactly. They used Instagram Stories polls to narrow down the undertone — warm, neutral, cool — and then TikTok comment threads to pick the finish. Matte, dewy, natural. By the time they went into production, they had a quarter of a million data points from their actual customers. Luna: That's wild. Traditional focus groups might give you thirty people in a room for two hours. Lucas: And those thirty people are often recruited, incentivized, and self-selecting in a way that doesn't reflect your real audience. What this brand did was essentially turn their social community into a product development department. And it's not just cosmetics. I'm seeing this in food, in apparel, even in software as a service. Luna: Is there a name for this? I've heard 'co-creation' thrown around, but that feels broader. Lucas: Some people call it 'community-led product development.' The idea is that instead of asking 'what should we build?' in a vacuum, you let the community signal what they actually want, and then you build it. The brand I mentioned didn't just ask once — they had a three-phase process. Phase one was the undertone poll on Instagram. Phase two was a TikTok series where they showed three finish prototypes and asked people to comment which one they'd buy. Phase three was a beta test where they sent samples to the top fifty commenters and asked for video reviews. Luna: So the community was involved from concept to near-final product. That must have built serious anticipation for the launch. Lucas: Huge. They had two million impressions before they even ran a paid ad. The launch post on Instagram got 40,000 saves in the first hour. People felt ownership because they'd literally voted on it. And here's the number that really got my attention: the brand told me their R&D cycle for that shade was about 40 percent shorter than their typical new product development timeline. Luna: Forty percent? That's massive. But I'm wondering — does this approach work for every brand? If you're a small business with 500 followers, you're not getting statistically significant data from polls. Lucas: That's a fair point. The scale helps. But I think the principle scales down too. Even with 500 followers, you can run a poll and get 50 responses. That's 50 people telling you something. Compare that to a focus group of 8. And if you're really small, you can do it in a private Facebook group or a Discord server where engagement is higher. The key is that the feedback loop has to be genuine — you can't just ask and then ignore the results. Luna: Right, because the worst thing you can do is ask for input and then launch something completely different without explanation. Lucas: Exactly. That erodes trust fast. The brands that do this well are transparent about how they used the feedback. They might say '83 percent of you voted for warm undertone, so we went with that. Now help us pick the finish.' It becomes a conversation, not a survey. Luna: I want to talk about the risks, because there's a flip side to this. What if the vocal minority skews your data? A lot of brands have superfans who are not representative of the broader market. Lucas: That's a real danger. The loudest voices on social media are often the most extreme. So smart brands don't just rely on engagement metrics — they also look at purchase data, return rates, and repeat purchase behavior. The cosmetics brand I mentioned cross-referenced their poll results with actual sales data from their existing shade range to make sure the new shade wouldn't cannibalize their own products. Luna: So it's not 'community says jump, we jump.' It's 'community gives us a signal, we validate it with our own data.' Lucas: Exactly. And there's another risk: if you do this too often, you can fatigue your audience. Nobody wants to vote on every single product attribute. The brands that do it well reserve community-led development for the products where customer preference really matters — like shade, flavor, or design — and keep the technical decisions in-house. Luna: That makes sense. Let's talk about the platform strategy. Instagram polls are great for quick binary choices, but TikTok comments can be messy. How do you structure the feedback so it's actionable? Lucas: Before I get into that — and this is totally related to what we're talking about — I want to mention something. We do these episodes every day, and we deliberately don't run ads. No sponsors, no commercials. That's a choice we made because we think the conversation is better when it's not interrupted. If that choice resonates with you, and you've gotten value from these episodes, you can support the show at buy me a coffee dot com slash fexingo. It's a simple way to keep this ad-free. Luna: Yeah, and honestly, it makes a difference. We hear from listeners who say they appreciate not having a mid-roll ad break right when things get interesting. So if that's you, that link is there. Lucas: Okay, back to platform strategy. For structured feedback, Instagram's poll sticker is actually pretty good for two-option choices. But if you want more nuance, you can use the question sticker and then aggregate responses manually. Or you can use a third-party tool that integrates with Instagram to run ranked-choice polls. Luna: TikTok is harder because comments are public and you get a lot of noise. But some brands are using the 'add yours' sticker or creating a duet challenge where people show their preference. Lucas: Right. And for more in-depth feedback, a lot of brands are moving to private communities on Discord or Circle. That's where you can have threaded discussions, run polls, and even share prototypes. The cosmetics brand I mentioned actually invited their top 100 commenters to a private Discord server for the beta test. That gave them detailed feedback without the noise of a public feed. Luna: So the public social channels are for broad signal detection, and the private communities are for deep engagement. That's a smart two-tier system. Lucas: Exactly. And the brands that do this well also close the loop publicly. They'll post 'you voted, we built it, here it is' and tag people who participated. That turns the launch into a celebration of the community, not just a product drop. Luna: I think we'll see more of this in the next year. Brands are hungry for faster, cheaper market research, and social media gives them direct access to consumers who are already engaged. Lucas: For sure. And as AI tools get better at analyzing unstructured feedback — like comment threads and video transcripts — the barrier to doing this at scale will drop even further. A brand might be able to run a TikTok campaign, feed all the comments into an AI model, and get a ranked list of product attributes in hours instead of weeks. Luna: That's the future. But for now, the playbook is pretty clear: start with a specific question, pick the right platform for the kind of feedback you need, be transparent about how you'll use it, and close the loop by showing the community what they helped create. Lucas: And one last thing: don't forget to thank them. The cosmetics brand sent a personalized thank-you video to everyone who participated in the beta. That kind of gesture turns a one-time participant into a lifelong advocate. Luna: Alright, I'm inspired. I might run a poll on our own social channels to see what our listeners want to hear next. Lucas: Now you're thinking like a community-led brand. Just make sure you actually follow through on what they say. Luna: Deal. Thanks, Lucas. Lucas: Thank you, Luna. And thank you for listening.