Digital Clubhouse April: The Customer You Think You Know
Tony Conte
Founder
April’s Digital Clubhouse was one of those events that reminds you why the whole thing exists in the first place. It was useful. Honest. A little messy in the way real business always is. And, crucially, full of people who are actually trying to build things rather than simply talking about building things.
That distinction matters. The industry is full of people who can describe a market but never really have to fight in it. This wasn’t that. Ryan Murray, Harriet Noy, Guillaume and Ben from Opteo, and James Sturdy all came armed with stories that only exist when theory collides with customers, budgets, operations, and all the tiny humiliations that come with trying to make commerce work.
I hosted the panel on the chameleon customer, which is a phrase I wish I’d invented, but didn’t. It’s too good not to use. It captures something that feels obvious once said out loud and strangely elusive before that: people do not stay in the box we build for them. They wander. They contradict themselves. They say one thing on Tuesday and do another by Friday. They are, in other words, human. Marketing has spent years pretending otherwise. That habit is expensive.
Ryan started his career at Marks and Spencer, moved into lingerie buying, helped ASOS get going, and then bought Mimi, a luxury lingerie brand, out of administration in 2019. That alone would have been enough drama to last a lifetime. But the detail that makes it interesting is that he did it with the world about to shut down around him, stock stranded in China, and no obvious reason to believe the timing would be kind. He bought anyway.
That’s what separates operators from spectators. Spectators wait for clarity. Operators make calls in conditions that are usually imperfect and occasionally ridiculous.
The smartest thing Ryan said was about objectivity. Running a brand outside your own demographic strips away vanity. It forces you to listen to what the customer is telling you rather than what your own taste would prefer. And when you hear that more than 60% of Mimi’s online customers have bought again, while returns sit at 12% in a category where 40% to 50% is normal, you are not looking at a marketing miracle. You are looking at a product doing its job properly. Fit matters. A lot.
Ryan’s CRM point was just as revealing. When the business relaunched, they initially did the thing many brands do when they have something to announce: they announced it. Broadly. To everyone. “We’re back. Here are the new styles.” That is not segmentation. They fixed it by getting serious about the database — one email per customer per week, but with twelve versions depending on where each person sits in the journey. That is the sort of thing that sounds boring until you realise boring is often where the money lives.
Harriet Noy’s talk had the energy of a founder still in the middle of the climb. That always makes a session better. Finished stories can be polished into submission. Active ones carry friction, doubt, and momentum. Hazaar began as a student marketplace and now operates as a sharper, more deliberate way for brands to connect with students. That shift tells you a lot about where the real opportunity sits.
Most student marketing is lazy. It relies on discounts, codes and a vague hope that being young on the internet counts as a strategy. Harriet’s model has more intelligence than that. It recognises that students are not a single behaviour block. They are price aware, yes. They are also identity aware. They want brands that feel useful, desirable and socially legible. A student does not simply want cheap. A student wants something that feels like a smart choice.
Her examples showed how that works in practice. A running shoe activation at Cardiff Met. An Oh Polly Halloween event where students collected returns as costumes. A giant block of ice on Nottingham campus for a Peak Performance launch. These activations create a world around the brand. They create a place where the sale can happen with some texture and some memory attached to it. That is much harder than publishing a discount and waiting for conversion.
The relaunch story was the most revealing part. Harriet stripped the platform back and curated the brands around what students actually want, including Birkenstocks, Uggs and Hollister. Sensible. Discovery only works when the consumer feels the pull of recognition. If the shelf looks random, the model collapses. She was also brutally honest about the mistakes along the way, including the wrong platform, the wrong funding direction and a friendship that broke down under the weight of an equity dispute. That is painful, but it is also real. The business world runs on these lessons whether people admit it or not.
Opteo’s session gave us the clearest reminder that technical sophistication and commercial clarity are not the same thing. Guillaume founded the business thirteen years ago as a Google Ads consultancy, built the software with his best friend, bootstrapped the whole thing, and then spent the last 18 months rebuilding the platform from scratch so it could become properly multi-channel.
The scale matters here. Opteo influences around £1.5 billion of UK ad spend. When a business sits at that level, it sees the market in bulk. It sees what happens when automation spreads, when channels blur, and when marketers start mistaking platform convenience for strategic progress. Guillaume’s line that paid search looks easier than ever but is actually harder was one of the best of the day. It is plain, accurate and deeply inconvenient. Which usually means it is right.
Platforms have spent years packaging complexity into interfaces that look simple. That creates the illusion of mastery. It flatters the user. It also creates room for sloppy thinking. If the machine does more of the work, the human has to understand more of what the machine is doing. The better the automation, the more important judgement becomes. That should be obvious. In practice, it is still widely ignored.
Ben’s point on AI and trust was the right note to end on. AI can produce useful work very quickly. It can also produce polished nonsense with absolute confidence. That is the danger. Feed it the wrong conversion action or the wrong date range and it will give you a report that looks decent and means very little. In a commercial environment, that is a serious risk.
Interested in Brave but not quite ready to speak to us? Stay in the loop.
The panel began with the idea of the chameleon customer: the shopper who refuses to stay in the tidy box marketing built for them. They search on Google, scroll TikTok, ask AI, compare on marketplaces, check paid ads, read organic results, message friends, and probably make the actual decision in a WhatsApp chat five minutes before checkout. They are premium on payday, careful by midweek, loyal until a better option appears, and then not so loyal at all. Which is to say, they are human.
That human-first thinking is what so many businesses fail to plan for. We like personas because they are neat. We like journeys because they are drawable. We like funnels because they look like something you could explain in a meeting without sweating. But actual customers are not diagrams. They are shifting, contradictory, context-driven people with budgets, moods and habits that change more often than our campaigns do.
James Sturdy made one of the most useful points of the afternoon: the persona is not dead, but it is incomplete. That is exactly right. It should be the beginning of thinking, not the end of it. His business runs weekly insights meetings with marketing, customer service and leadership together. That sounds like common sense because it is common sense. Most organisations fail at it because common sense requires coordination, and coordination takes effort. One of the insights James shared was a lovely example of the value of staying close to the customer. They discovered that a significant number of people were replacing the traditional Sunday roast with a pie. Cost management, not culinary evolution. The type of insight that will never appear unless someone has done the work properly.
Guillaume added the helpful corrective that qualitative data still matters. More than that, it matters precisely because quantitative data can only carry you so far. Numbers tell you what happened. People tell you why. And if you want to understand behaviour, not just describe it, you need both. Watch people use the product. Listen to how they explain themselves. Notice where they hesitate, over-explain, or contradict their own logic. That is where the truth usually lives.
Guillaume’s take on AI was an investing one; describing it as a yes-machine, which is wonderfully blunt and probably accurate. It smooths, agrees, and reduces friction. That sounds efficient. It also risks creating a generation of consumers who become even less tolerant of challenge, delay, or disagreement. Ben’s point was the complement to that: AI is making it easier for more things to look good. That means trust, real trust, becomes more valuable, because surface quality is becoming cheap.
The best thing about the day was that it felt grounded. Ryan reminded us that product quality and segmentation matter more than vanity. Harriet showed that curation and community beat blunt discounting. Opteo showed that automation changes the shape of the challenge, not the need for human judgement. James showed that real customer insight comes from proximity, not assumption.
I think the danger is that Brands are often seduced by their own language. They confuse activity with progress. They trust dashboards more than customers. They treat the market like a spreadsheet and then act shocked when people do something messy and inconvenient.
The final thought is simple. The brands that will keep winning are the ones that get closer to customers, build things worth returning to, and stop confusing polish with substance. That is where the money is. And that is where the work should be.
