1/4 - How can artificial intelligence be useful to fashion? The answers, supported by concrete examples, in this first part of the interview of Romain Chaumais, executive leader of the French company Fashion Data, specializing in AI applied to fashion.
By Ludmilla Intravaia
Le Boudoir Numérique : Fashion Data is a French start-up of artificial intelligence dedicated to fashion, created in 2019. What is its mission?
Romain Chaumais, executive leader of Fashion Data: Fashion Data aims to help the fashion industry transform towards a new zero waste fashion business, using data and artificial intelligence technologies. The fashion sector is the second most polluting industry in the world, responsible for 20% of wastewater discharges and 10% of carbon dioxide emissions. It cannot last any longer: the textile industry must radically reduce its environmental footprint. This sector must also face a major change in consumption habits, with customers now aspiring to a much more reasonable and sustainable fashion. The consumers are more likely to review their appreciation of the brands they buy and their behaviors are constantly changing. We are also witnessing an evolution in the path-to-purchase which is becoming more and more digital, which requires revisiting the role of stores, economically threatened. Today, they represent less and less the major interaction in the act of purchase, they intervene rather at its beginning or at its end. However, stores are essential to preserving jobs. In France, 180,000 people work in fashion stores. It is essential to maintain the economic role of the store, but also its societal role, because shopping has been part of our European culture for a long time. It is hard to imagine fashion only living in digital. Finally, with the textile and clothing market having lost 15% of its value in recent years, our mission is to support an industry in degrowth which, at the same time, must invest in respecting the environment. We often talk about eco-responsibility. But that is not enough. Fashion companies must remain profitable if they are to continue to innovate and reinvent themselves. This is why I prefer to talk about eco-profitability. And not just for fashion. All industries in the world, automobile, aviation, tourism for example, will have to follow very quickly this path to do better with less. This is the direction taken by history.
How do you help fashion companies achieve eco-profitability?
By putting data and artificial intelligence at the service of three business organization streams. First, the customer. The better we know customers, the better we support them, the more we strengthen their loyalty and engage them with the brand. Second, the product or how, by better detecting trends, we can produce a collection that meets consumer expectations and adjust the quantities to manufacture only what a brand is able to sell. The first step towards ecological fashion is not to manufacture products that will be unsold tomorrow. By avoiding the over-manufacture of unnecessary products, we are maximizing our action in terms of environmental footprint. The optimization of logistics and the store are brought together in the third stream on which we are working because one does not go without the other. Many ready-to-wear brands suffer from a lack of intelligence in allocating products, in the right stores, at the right time. The performance of a store is of course linked to its culture of animation by salesmen, but also the efficiency of its supply chain.
Your first lever of action is the customer. Why is it important to know him as well as possible?
Each of us is driven by a strong desire to be recognized for who we are, for what we love, and for all of the interactions we've had with the brand. Coming to the same store fifty times and always being seen as a stranger is upsetting. On the contrary, nothing is nicer than entering a store, where you come often and being greeted by a salesman who will recommend you these new pants that would go so well with the blouse you bought, during your previous visit and give you immediately, in the fitting room, the right size, because, quite simply, he knows you. This service and commercial relationship that good salespeople provide in the store is very flattering and we expect it from others in our path-to-purchase. In the digital world, via a website, via the e-mail or the SMS that you receive, the advertising brochure in your mailbox, you have this same aspiration to be recognized for who you are. Nothing is more annoying than seeing a T-shirt in a brand newsletter with 25% off when you bought it at the full price, or being suggested items in sizes that are not yours. So this customer experience, which is super natural in the real world but which, in the digital world, is driven by a machine, must be reproduced.
With artificial intelligence?
Absolutely. Our algorithms are intended to offer the same comfort to the customer who is recognized, pampered, as well advised in the digital world as he is in real life. And since we can't put a salesperson behind every Internet user on a website, we put algorithms and artificial intelligence that will do it for them, on a very large scale. This AI and these software will mimic this subtlety in the service and the right recommendation, what is more, simultaneously, on all the interaction channels that you have with the brand, on the website, in your e-mail, in the store, with the loyalty card, taking into account everything you may have done beforehand, in terms of navigation, purchase, etc. In fact, knowing the customers better allows us to serve them better, in a digital universe that is much colder, much more brutal than the real world, thanks to all these algorithms reproducing the quality of service that we know in stores, when we interact with good salespeople.
Can better knowledge of the customer through artificial intelligence also be useful for fashion collection managers, for example?
If a collection manager or a stylist can see everything that is sold, everything that is tried on in the fitting room but never bought, that these pants are never worn with this sweater, when he thought they would be, that this outfit is selling well with one type of clientele and not at all with another, in short if he has a better understanding of the client and of the way his tastes evolve, he can adapt with product adjustments, revisited sizes, capsule collections, to better meet the consumer expectations. If you’re the loyal customer of a shop that always serves you the same thing, at some point you get bored. And since fashion must perpetually reinvent itself, knowing the customer makes it possible to understand, beyond the quantities sold, why, how and by whom they were purchased. Studies of panels, representative of consumers, have always existed in stores, of course. But that's just a sample. With big data, we are not going to process a sample, we are going to take all the sales of all customers, all the information we have on them, to cross them with other external data, from open data in free access to all, for instance, such as meteorological information or socio-demographic profiles and thus be able to observe in detail all the phenomena that a single person is not able to analyze on his own.
Does artificial intelligence have a role to play in understanding the omnichannel customer?
The omnichannel life cycle is a major issue for all traditional retailers. If we look at the pure players of the web, the Digitally Native Vertical Brands (DNVB) which only sell online, their job is ultimately quite simple. Basically, you have a warehouse and a website, possibly coupled with a mobile application, which is your sales channel. People buy from your site and you deliver the items from your warehouse. If the people are not happy with the product, they send it back to you there. With the omnichannel customers buying from your website and in the store, everything becomes much more complicated, since you must simultaneously offer the same products online and in stores, at the same prices. When you have placed the items in your store, they are no longer available in your warehouse and when you make a sale on your website, if you no longer have any products in your warehouse, you must send it from your store, using ship-from-store strategy. You have a lot more constraints, because the customer journeys are very different. The consumer can start shopping online, then go to the store and return to finish the purchase on the internet. Or receive an email, browse the web and buy in a store. Buy online and then return to the store. Buy in a store and finally change his mind and take back the product to another store. Artificial intelligence helps support consumers in their omnichannel path-to-purchase and stimulate the use of all sales channels. Piloting the omnichannel life cycle aims to push the customer to buy simultaneously on your website and in your store, when a consumer only buys online, to invite him to do the same in the store and when he no longer buys in stores, to make sure he chooses the web. Fostering the omnichannel path-to-purchase is fundamental for loyalty because when a customer knows that the product is available online and in stores, he has the reflex wherever he is, at work, on his sofa with a tablet or on the move with his phone, to come and see you either digitally or physically, with the conviction that he will get all the answers to his wishes. This omnichannel customer is said to have 30% added value than other customers.
* Continue reading on artificial intelligence applied to fashion, with the second part of Romain Chaumais' interview on Le Boudoir Numérique: "Big data enables intelligent management of the fashion stores supply chain".
* Fashion data website is here.
* Another papers on Le Boudoir Numérique :
- What will we wear post-lockdown ? The answer with Heuritech’s AI
- Covid-19 – Heuritech launches a solidarity program for fashion brands
- “Cross-fertilization between tech and fashion is the strength of Heuritech”
- “To help fashion brands decision-making with AI during the Covid-19 crisis”
- Will leopard print pleated skirts be trending any time soon ? AI already knows it !
- “AI can contribute to the virtuous circle of sustainable fashion”
- “Our algorithm is the link between the parfumer and the customer”