Does data add great value in the fashion world too? Or does it hinder the creativity of designers and their businesses? We asked data experts Philippe Baecke and Michael Lemner, who both see a lot of untapped potential for fashion companies, far beyond marketing. “But the creation process will always be about balancing data and intuition.”
© Paulien Verheyen
Wil je op de hoogte gehouden worden van nieuwe artikels in dit magazine? Schrijf je in op onze nieuwsbrief!
Philippe Baecke: “Data makes business more predictable, but fashion is singularly unpredictable”
Philippe Baecke is a Full Professor of Business Analytics & AI at the Vlerick Business School. Compared to other sectors, the fashion sector is not yet particularly mature in that domain, he notes. “The focus in the fashion world is very much on creativity and intuition, which instils a sense of reticence. This reticence does make a certain sense. You use data to recognise patterns, to make your business more predictable. But fashion is singularly unpredictable: trends come and go with the seasons.”
“Does that mean data is worthless in the fashion world? Of course not. Fashion entrepreneurs are already making plenty of use of data for marketing purposes, but there is still untapped potential. I am thinking, for example, of demand forecasting, one of the biggest challenges for the fashion sector. With collections changing four times a year, is data from last year still relevant to predict demand for this year? I think so. You can use image recognition based on AI to assign a similarity score between products across seasons. So you can better predict demand, manage stock more efficiently, and reduce the risk of shortages or large overstocks.”
The rise of AI
Philippe Baecke does not think big fashion companies with lots of financial and technological capabilities have a leg up when it comes to data. Quite the contrary, in fact. “Research has been done on the factors that determine data maturity. A company’s size was found to play only a limited role; C-level support and company culture were much more decisive.”
“Fifteen years ago, small businesses with limited resources had a harder time, in terms of IT; today, there are countless cloud and SaaS solutions that make data collection and analysis much more accessible and approachable. Of course, you cannot ignore the advance of Artificial Intelligence in this regard. AI cannot do without data, but it also makes it possible to do much more with that data.”
“I distinguish between different phases of AI and data maturity. In the final, most mature phase, companies go from scratch to develop AI solutions themselves. I honestly don’t think there is one Belgian fashion company which does that today; that’s really next level. In the first phase, you can just use what is available on the market. Tools like ChatGPT and Midjourney are trained on millions of data sources, which already provide very valuable info and insights that companies can work with.”
“At an intermediate stage, you can also adapt existing AI solutions to your business. For example, you could use customer contacts, via email or other channels, to find out how satisfied your customers are. There are models for this that dissect text conversations and extract sentiments. You can then fine-tune those models to tailor language that’s specific to the fashion industry.”
“If you focus on the data sources that are really useful, then you can quickly realise 80% of your use cases by bringing 20% together in one database.”
What advice does Baecke have for fashion entrepreneurs who are still lagging behind on data? “Start working out a data strategy. That sounds bombastic, but it means you should look around your organisation and identify x number of use cases where data can add value. Look beyond the short-term; set the horizon at one to three years. Or even five years. You’ll see that those use cases also overlap in terms of data and analytical requirements.”
“The biggest challenge is not collecting data, but bringing together data from different sources. Data today is often scattered across CRM systems, ERP systems, sales teams’ Excel files, social media, etc. That seems like a real mountain to cross, but by no means is all the data equally relevant. If you focus on the data sources that are really useful, then you can quickly realise 80% of your use cases by bringing 20% together in one database.”
Michael Lemner: “Fashion entrepreneurs will always have to balance data and intuition”
Michael Lemner is a true fashion veteran. He launched H&M in France and in Belgium, he managed LolaLiza, and was active in several other companies. Now with Tim-Tam Consulting, he focuses on the importance of data for the retail world, including emerging fashion tech start-ups.
“The challenge for fashion companies has changed radically,” says Lemner. “They used to wonder where on earth to find that data. Today, in the digital age of online shopping and social media, the main question is how to select qualitative and relevant data from the avalanche of information at your disposal. And in a second phase, how to extract the right insights and actions from that relevant data.”“Small and medium-sized players still too often stare blindly at marketing, while data can also help solve logistics problems, for example.”
“You have roughly two types of data: product data and consumer data. Fashion companies know the product data; they are simply in the label and on the price tag of each piece. Where was it made? What is it made of? How much does it cost? Today’s fashion companies also already have a wealth of consumer data. They have a very good view of who bought a piece and how consumers came to buy it, especially when it comes to online shopping. But even offline, loyalty cards, loyalty programmes, and discount codes already give a pretty revealing insight into exactly who that consumer is.”
“It’s just that fashion companies sometimes still don’t do enough with data. I often come into shops and ask, for example, which colour sells best. The answer is often, ‘Uh, no idea. We’d have to check that out’. Certainly, the small and medium-sized players still have a ways to go. They still too often focus blindly on marketing, when data can also help solve logistical problems, for example.”
Customers and walk-outs
According to Lemner, an effective data strategy requires a clear focus. “I look at data with a retail lens from my own background and experience. The essence of retail has not and will not change: buying and selling. It’s especially crucial for fashion companies that do not have time, people, and resources in surplus to always keep that in mind when collecting data. Who are your suppliers? Who are your customers? What do they want, today and tomorrow? Focus on the data that gives you relevant insights into that; forget the rest.”
“There are a lot of effective, affordable tools available today to collect and analyse data. Don’t overcomplicate it; don’t invest money in expensive and complex technology if the basics are not right.”
Lemner has another striking piece of advice for fashion companies: collect relevant data not only about your customers, but also about the people who look around but leave again without buying anything. “Their data is not only as valuable as that of your customers, they are also much more so. We know from research that the average conversion rate is somewhere between 15 and 20%. That is, five out of six people who come into your shop walk out again or leave your website without actually making a purchase. Were they unable to find their size? Was that one piece they were looking for not there in the right colour? You don’t know.”
“I myself am active in a start-up, LiveCrew, which identifies who the ‘walk-outs’ are. Vendors usually do talk to those customers and often know why they end up not buying anything anyway. We make it very easy for those sellers to register that quickly and efficiently. Smaller and medium-sized brands have a trump card here, by the way; they have much more contact with (potential) customers than the big chains.”
Through the data jungle
Does Lemner also see a role for data in the creation process? “Call me old fashioned, but I still believe that fashion is ultimately about the clothes. The product is king, despite all the marketing around it. Designers will never rely on data alone when creating that product. They travel, they go to shows, they read magazines, they keep their eyes open on the streets. Data can help designers continue to see the forest – the jungle of data, so to speak – for the trees. It will always be about balancing data and intuition.”
