Latest / The DTC Marketing Podcast with Fexingo: Direct-to-Consumer Brands, Shopify, and E-Commerce Growth / How DTC Brands Use Customer Reviews for Product Innovation
Transcript
- Lucas: You know how every DTC brand puts 'real customer reviews' on its homepage, usually with a five-star average and a smiling photo. Luna: Yeah, standard social proof play. Lucas: Right. But a handful of brands are using those same reviews not just to convince new buyers, but to decide what to build next. I'm talking about treating the review section as a continuous, unpaid focus group. Luna: So instead of just displaying reviews, they're actually mining them for product insights. Lucas: Exactly. And the brand I want to anchor this on is Blueland — the cleaning products company that sells tablets and reusable bottles. They have a pretty vocal customer base, and they've built a systematic process to turn that noise into actionable R&D data. Luna: I've seen their reviews. People are passionate about the glass cleaner scent. Lucas: Passionate and specific. Blueland's team tags every incoming review by sentiment and by product attribute — things like 'scent strength,' 'bottle design,' 'tablet dissolving speed.' Then they run quarterly reports that highlight the most frequently mentioned pain points and requests. Luna: So they're not just reading the five-star raves. They're looking at the three-star complaints and the feature requests buried in four-star reviews. Lucas: Exactly. And the result was a reformulation of their dish soap. Customers kept saying the original formula left a residue on glassware. Blueland's team cross-referenced that complaint across hundreds of reviews, confirmed it wasn't a one-off, and tweaked the surfactant blend. Luna: That's a pretty direct line from 'someone typed a complaint' to 'we changed the chemistry.' Lucas: And it didn't stop there. They also launched a new scent — eucalyptus mint — because they saw a cluster of reviews asking for something more 'masculine' or 'earthy.' That's a product line decision that came straight from review text. Luna: I love that. But I also wonder — doesn't this create a risk of overcorrecting based on a vocal minority? Like, if you only listen to people who bother to write reviews, you might miss what the silent majority wants. Lucas: That's a real tension. And it's exactly what happened with a DTC mattress brand — I won't name them, but they leaned hard into review data to decide on a new firmness level. A vocal group of back-pain sufferers was clamoring for a firmer mattress, so they launched an 'extra firm' version. Luna: Let me guess — it flopped. Lucas: Pretty much. The extra firm mattress had a high return rate because most customers actually preferred the original medium firmness. The reviewers were a self-selected group with a specific problem, not representative of the broader market. Luna: So how do you do it right? How does Blueland avoid that trap? Lucas: They have a few safeguards. First, they only act on patterns, not individual requests. A feature request needs to appear in at least five percent of reviews for a given product before they escalate it to product team. Luna: Five percent threshold — that's a concrete number. Lucas: Second, they separate sentiment tagging from product decisions. The marketing team does the initial tagging, but the product team does its own validation — sometimes through surveys sent to a random sample of recent buyers. So they check whether the review feedback matches broader customer sentiment. Luna: That's smart. So the reviews are a signal, not the final answer. Lucas: Right. And third, they use review data to inform incremental changes, not wholesale pivots. The dish soap reformulation was a tweak to an existing formula, not a whole new product line. The eucalyptus mint scent was a line extension, not a rebrand. Luna: Makes sense. Smaller bets, validated by review patterns. Lucas: Exactly. And there's another layer here — some DTC brands are using AI to scale this process. Instead of a human tagging each review, they use natural language processing to automatically categorize feedback by theme and sentiment. Luna: I've heard of tools like ReviewMeta and Yotpo doing that. Lucas: Yotpo has a feature called 'Insights' that does exactly that. It scans all your reviews and surfaces top topics — like 'shipping delay' or 'fit issue' — with sentiment breakdowns. A brand like Olipop, the soda company, uses that to track how formulation changes affect customer perception over time. Luna: So they can see if a new sweetener blend is generating more 'tastes great' mentions and fewer 'too sweet' complaints. Lucas: Exactly. And that's the holy grail — closing the loop between customer feedback and product iteration. But it requires a culture where product teams actually read reviews, not just marketing teams. Luna: That's a big cultural shift for a lot of companies. Traditionally, product teams look at sales data and maybe NPS scores, but reviews are often seen as a marketing asset. Lucas: Blueland's head of product told me that they have a Slack channel called 'review gold' where anyone in the company can post interesting review excerpts. It's not formal, but it creates a habit of listening. Luna: I like that. Low friction, high awareness. Lucas: And it pays off. One review mentioned that the starter kit's bottle cap was hard to twist for people with arthritis. That went straight to the packaging team, and they redesigned the cap with a larger grip. Luna: That's a real accessibility win that also broadens the market. So the review section becomes a product development tool, not just a sales tool. Lucas: Exactly. And if that kind of thinking has ever sparked something you've actually used in your own work or business, honestly, if today was worth a coffee to you, there's a link — buy me a coffee dot com slash fexingo. It's how we keep this show ad-free and independent. Luna: Yeah, it's a small way to say this kind of conversation matters. Appreciate anyone who does. Lucas: So to wrap this up — the key takeaway is that customer reviews are underused as a product innovation signal. The best DTC brands treat them as a continuous, unpaid focus group, but they also validate the patterns before acting. Set a threshold, separate signal from noise, and use it for incremental improvements. Luna: And make sure product teams are reading them, not just marketing. That's the cultural piece. Lucas: Right. So next time you're scrolling through reviews for your own brand, ask yourself: what are customers actually asking for here? The answer might be your next product launch.