As much as we think of fashion as something intuitive, it has always required a great deal of knowledge and information. From the trickle-down and bubble up of fashion trends, to market statistics, and customers sizing details, it is all data which needs to be ethically and fairly processed to adhere to issues concerning privacy.
What is the role of the designer in this process? How should we handle this data for creative purposes, not only marketing and sales? Are there possibilities for co-creation using data, or utilizing this technology to engage with the consumer's real time feedback? Is data a friend or foe for the fashion designer?
We invited an expert group to answer these and other questions concerning how data can be employed by the fashion designer and the fashion system to optimize and streamline every part of the fashion supply chain.
Trust your instincts and your data
"You can say my gut feeling is backed-up by very early signals data", said Julie Pont, Fashion & Creative Director at Heuritech. The Paris-based startup is a trend forecasting solution for fashion and luxury powered by Artificial Intelligence. Heuritech uses machine learning algorithms to trawl millions of social media images and analyze, with the help of their fashion team, early trend signals and how they are operating in different markets.
Traditionally trained as a fashion designer, Pont commented on how it wasn't easy to trust data at first. "I was a bit scared, to be honest, due to my creative background. Could this replace my job? I was intimidated by what could be the added value of data. I was operating on gut feeling, like we do in fashion. As a designer, you know that your instincts and your creativity are the reasons why you've been hired. But I soon realized data is not here to replace designers but to help with the uncertainties." adds the Parisian fashion designer.
Pont sees data as a time saver for fashion designers. "As a creative person there are many sources of inspiration you can research but you have so much demand from your customers, from the brand, and your market. Technologies such as AI powered trend research can optimize the time and quality of the designer’s research and improve the decision-making processes. To me, as a fashion designer, I rely on data to secure my creative decisions. It validates my ideas” states Pont.
Data analysis and consumer behaviour
If from the designer's perspective data can help to make better decisions, from the consumer side, this is not always the case. For Jonathan Chippindale, Founder & CEO of Holition, a creative technology agency in London that creates immersive experiences for retail, the secret to analyse data correctly is to look at consumer behaviour and understand that it can always change.
"There is a massive sense of irony with algorithms. We have infinite information and data that we have access to, but algorithms are funneling it down into areas where it thinks we might be more interested and removing areas that are less interesting to us. In some ways, it is pushing us into the middle of the bell curve where you take all the colors, mix them all together, and you end up with grey. It's the average. But that's not human, is it? That's not the way we are. We all dress differently. We will behave differently. We'll talk differently. We all like different things," says Chippindale.
And he adds: "If we can recognize where the emotion comes in, if we can recognize behaviour, but also the causes of behaviour and feed all of that into it, that's interesting to me. Recognize a propensity to take risk and a propensity to discover. ‘I want to go on to that mountain top that because the view is going to be great. No one's been up there before, but I'm going to go up there.' That's what digital is all about."
For Chippindale, the impact of digital in the fashion industry today is particularly focused on how brands deal with the consumer: "brands have had to let go of control. When I was a marketing director, brands used to tell women what to wear, how to wear it and when to wear it. That behaviour just seems crass today, someone telling you what to wear, that just feels wrong."
More data, better products?
If telling the consumer what to wear feels wrong, how can brands position themselves when they have access to consumer data and can influence purchasing decisions? The question is about consent, and if I as a consumer, agree to the brand gathering my data and developing better products for my use, the control is with me.
But how can a company make consumer consent related to data viable and scalable? For Beth Esponnette, Co-Founder, Chief of Product and Executive Chairman at Unspun, a robotics and digital apparel company building custom-made jeans, the answer is an on-demand system. "The more data, the more information you can get on that person, the more intentional the product is going to be." This is a key concept for Unspun, as Esponnette adds: "We are attempting to flip and start with the customer and then produce a product for them. The customer takes a body scan, chooses the design they want, and then we make the product for them."
And does the customer always know exactly what they want? "We have to sift through that", says Esponnette, "If they tell us, 'No, I like to wear my jeans like this', they don't always know exactly how, but this is something that we collected over time and we put into our algorithms. But that leads to two issues: are we working bias? Is there bias in this based on how we've categorized people?" questions Esponnette. The more we research how data can impact the fashion system, the more complex the issue presents itself, making it even more important to know with whom we are sharing our data.
Data can be used by fashion designers to better know their customers, to help them in their creative process, to make smarter decisions based on real-time trends, and/or even choosing a different aesthetic option. The opportunities are endless if we respect data and where it comes from. Data is generated by our behaviour as consumers online and offline. All we do, everyday generates data that can help us, or be used against us. The issue is to use it and share it wisely. As Jonathan Chippindale added at end of his interview, "the algorithm is not an oracle, we are the ones who should be questioning.”