The Beauty Data Gap (And How We’re Fixing It)
Barefaced has pivoted into software, and this is my ambitious plan to solve beauty’s biggest blind spot.
Hey! Long time no chat — sorry for the radio silence. Barefaced has gone through quite a pivot, and I’ve also been deep in the trenches of uni exams. But I’m finally coming up for air, and I wanted to fill you in on why I’ve made the change, what we’re building now, and how it’s going to shape this Substack moving forward.
TL;DR: Beauty brands are trapped in a trend echo chamber, launching the same products over and over. I'm building Unfiltered to break the cycle. Unfiltered is a beauty review platform powered by real user data and demographic context, built to help consumers research better and brands build smarter. Moving forward, this Substack will share the insights I’m uncovering in the data. (Also, the podcast will be back next month.)
So today, I thought I would walk you through the entire business plan. This is exactly what I’ve been building and pitching behind the scenes, and I’m excited to share my plans to reshape how beauty products are both created and marketed.
Barefaced’s Thesis
When I started Barefaced, I set out to solve what I believed was beauty’s core problem: no one had the time to dig into the data. Terms like “data-driven” were popping up more and more in LinkedIn job descriptions, and everyone I spoke with in the industry understood how powerful data could be, yet there were hardly any data science or analyst roles within beauty companies.
My idea was to build a solution that could deliver those insights to brands without needing to hire someone. They knew they needed it, and if I could write custom code to scrape, analyse, and interpret the data, I could fill that gap. I would be doing work I love and work that they genuinely needed. And while I wasn’t wrong, these Substack reports gained traction quickly, I uncovered something even bigger: a growing demand for tailored insights.
Learning 1: Beauty brands don't want pre-packaged reports — they want consultants.
Publishing these reports quickly opened the door to a wave of unexpected opportunities. Beauty founders and marketers began reaching out for consulting, many of them weren’t even paying subscribers, just people who had seen the free previews of my Substack reports. Yet they were willing to invest thousands to apply that same level of insight to their own product strategy, marketing, and positioning. The demand kept growing, and suddenly I was getting full-time offers from major industry players, the kind of roles I would've jumped at six months earlier.
That was the point I knew I was on to something. I was right that the appetite for beauty data was strong, but the way I was delivering it wasn’t aligned with how brands wanted to use it. I also learned that outlining my methods of collecting the data (the hardest part of what I do) interested only those wanting to work in beauty tech, uploading the entire raw datasets interested barely anyone, but utilising my experience in the industry to translate the numbers into commercial insights was what beauty professionals wanted access to and would pay for.
Learning 2: The Beauty Data Available Is Limited
Being a full-time student meant I could only take on a handful of consulting projects, and I also wanted to prioritise building out the reports. Both experiences gave me the chance to operate as a beauty analyst, and more importantly, to understand what kind of data brands actually had access to.
What I found was that most of the tools were great, but almost all of them focused on search data and social media monitoring. They helped track trends, show what was gaining or losing traction, but in my mind, this was only phase one in measuring market interest. What I started to notice was that many brands relied entirely on search data for product development, and it showed. Everyone was launching the same products.
Brands were spotting spikes in interest around certain products or ingredients and responding by launching related products. Those new launches then fed into the trend, making the interest grow even more, which then signalled to other brands to join in. The cycle kept repeating. It’s turned the beauty industry into this trend echo chamber where we’re seeing the same products being launched again and again and again. And honestly, it makes sense. Before you create something, you need to measure it, and if search data is all you’ve got, then that’s what you build from.
This meant that for every client project and almost every report, I had to manually source an additional dataset to enrich the search data. I’d identify a growing category, then ask: Who’s driving this? How old are they? What’s their skin type, skin tone, nationality, and ethnicity? And in an ideal world, how have these demographics shifted over time? Take lip liner, for example. I knew that trend was created and nurtured by Black girls. British Black girls, I suspected. But I couldn’t prove it.
Despite the beauty industry constantly talking about becoming more personalised, there were no tools to support that shift. The power and influence of the communities shaping beauty were getting lost in the data, and everyone was losing out. These communities weren’t being served with the right products, and brands were leaving money on the table.
Learning 3: Beauty shoppers want these insights too
An insight I didn’t see coming was that beauty shoppers wanted this data too.
While building these reports, I started sharing some of the by-products of my research on TikTok. One example was a Google Sheet that catalogued Mecca’s entire holiday collection, ranked by best value savings and cost per millilitre. That sheet was used nearly 10,000 times by beauty shoppers who were clearly just as hungry for structured, comparative insights.
This unexpected traction revealed that beauty consumers aren’t the passive trend followers they’re made out to be; they’re more informed and more discerning than ever. And when I thought about it, this made sense. The way we discover beauty has shifted dramatically (think TikTok), as has the way we purchase it (hello, Amazon). But the crucial middle step, the part where we evaluate, compare, and figure out what’s actually right for us, remains completely fragmented. Finding raw, relevant reviews from people who look (and live) like you often means diving into Reddit rabbit holes, asking ChatGPT for ingredient breakdowns, and bouncing across countless websites just to work out what’s worth your money.
The need for a tool to help them cut through the noise became obvious.
The Pivot
All of this came together and sparked the idea for a two-sided model. At its core, Unfiltered is a database of demographic-rich beauty insights with two unique access points.
For consumers, it’s a review-based app that organises real experiences around the factors that actually impact product performance—like skin type, tone, age, ethnicity, nationality, accessibility needs and more. Because when someone asks, “Will this work for me?” they’re not looking for a brand story. They want honest insights from people that look like them.
For brands, Unfiltered provides structured, qualitative insights from real users. It helps identify unmet needs, guide thoughtful marketing, and support the development of products that truly resonate.
Of course, building a product like this doesn’t come without challenges, and some major questions to address.
The First Problem: A Review Platform With No Reviews
No user wants to sign up to a review platform with no reviews, and no brand wants access to an empty database, so I knew we needed to start with data from somewhere.
While major retailers and brands often publish product reviews on their own sites, beauty shoppers know these come with a grain of salt — at the very least, they’ve been reviewed by the brand. When I’m searching for honest insights for the brands I consult with, I always turn to the same place: Reddit. It’s home to countless thriving communities of beauty lovers, and with strict rules against self-promotion, it’s one of the few places you can consistently trust to hear what people really think.
But when it comes to actually finding reviews and conversations about a specific product or brand, it’s chaos. A single product might be mentioned in threads across a dozen different subreddits. You might care just as much about what olive-toned makeup artists say as you do about input from the sensitive skin community, but the platform simply isn’t designed to organise reviews in that way.
Imagine an encyclopedia of Reddit reviews, where for any given beauty product, you could instantly see which communities love it, which ones don’t, and exactly why. That’s version one of Unfiltered.
My life lately has revolved around developing an algorithm to collect and categorise hundreds of thousands of Reddit comments, sorted by demographic details and specific use cases. It’s been an undertaking more frustrating than I could ever properly articulate. But what it’s given us is a strong foundation: a database as deep and rich as a series of expensive focus groups, but spanning an entire category. (I’m starting with skincare — more on that soon.) It’s the kind of resource that both beauty nerds and beauty businesses would get a lot out of seeing. But this brings us to the second problem.
The Second Problem: Building Two Products at Once
A two-sided model allows us to do something powerful: help brands access and decode real reviews, while also building the Unfiltered app entirely around creating the most valuable and enjoyable experience possible for beauty lovers. No ads, no product placements, no “buy now” links, just real insights. But it means we’re effectively building two products at the same time.
At first, I thought, How hard can it be? The database is the hard part. Once that’s in place, surely we just need a simple dashboard so brands can view it. I was wrong. Like I mentioned earlier, raw data alone isn’t useful to most brands. The value lies in being able to query it, surface patterns, and extract commercial insights that can directly influence product development, marketing, and strategy. And for that to work, the dashboard has to be just as high-quality and intuitive as the data itself — otherwise, we completely undermine the product.
Monetisation
If we don’t have the capability of building a (great) SaaS tool but have this incredibly valuable dataset, how can it be utilised by brands? APIs.
From day one, a cleaned, categorised dataset of Reddit reviews, organised by an expansive range of demographics, is valuable to beauty’s biggest players. While not offering a dashboard limits the number of brands we can directly support, providing API access dramatically increases the value Unfiltered offers to those with the infrastructure and analytical teams to leverage the data.
Imagine Unfiltered data integrated directly into a brand’s existing Power BI, Tableau, or Qlik dashboards, right alongside sales figures and social listening tools. Suddenly, brands gain a new level of depth, with the ability to connect their metrics to real people and real-world experiences. Plus, no need to wait for a custom dashboard or report, just query the data however you need.
I’m not naïve about the challenge of going after enterprise customers from the outset. If we need to pivot, we will. But if we can pull this off, Unfiltered can focus on building the kind of trusted, no-BS platform that a beauty brand simply couldn’t create themselves (because of the inherent bias).
Growth
This structure would allow us to grow both a community and a proprietary review database, one that is far more valuable than anything we could ever scrape from Reddit. Reddit helps us understand who is using a product, but Unfiltered reviews allow us to understand how products are being used.
The goal of Unfiltered is to build a Letterboxd or Goodreads for beauty, a place where reviews are not a one-off but part of an evolving log of experience. For one skincare product, for example, you could log:
First impressions: Thoughts on the texture? Was it the packaging that sold you? Or was it a gift?
A few weeks later: How did your skin react? Are you using it in the AM or PM? How many times a week? Have you noticed any changes to your skin yet?
Three months in: Are you still using it? Will you repurchase it? If not, what did you switch to?
These are the kinds of insights beauty shoppers actually want. Because someone raving about a product after three days is not the same as someone who is on their seventh bottle. And brands need this depth too. They need real consumer journey intelligence so they can move beyond the TikTok trend echo chamber and start addressing the countless white spaces within beauty.
And that’s Unfiltered! A pretty mighty undertaking for one full-timer and two part-timers (more on the team later). But if anyone can pull it off, there is zero doubt in my mind that it’s us.
The other great thing is that now I have access to an incredibly rich database! This Substack will be the perfect place to share the kinds of insights you can expect to find in Unfiltered data. The first one will go live next week.
Until then, you can find Barefaced everywhere right here, and you can go and create your Unfiltered account too!
So cool can’t wait to see this come to life !! 🥂