Yoox looks seeks AI’s help for design inspiration

Source: Yoox
Dec 18, 2018
Matthew Stern

Designers of Yoox’s latest apparel line are getting a little inspirational help; the line is based on information collected, aggregated and powered by artificial intelligence.”

Yoox’s AI tools review content across social media and online magazines to furnish the retailer’s designers with data on what styles are being discussed and shared. Inspiration also comes from predictive indicators into emerging lifestyle and style trends, analysis of internal data from products sold on its site, customer feedback, industry purchasing trends, as well as text search and image recognition.

“We are always exploring the frontier between Human and Machine; it’s been part of our DNA since I invented Yoox in 1999,” said Federico Marchetti, CEO of Yoox Net-A-Porter Group, in a statement.

AI-driven apparel design, however, raises a question: Is success more dependent on being responsive to feedback on current trends or on pioneering new trends customers yet asking for?

Yoox has not jettisoned more analog modes of coming up with design ideas. A new nine-piece Mary Poppins-inspired line, tied to an upcoming movie release, offers contemporary upgrades on the character’s classic fashions, according to Bustle.

Others in the fashion world have used other implementations of AI to try to anticipate the kind of products customers are looking for. A startup called Choosy bases designs on data from Instagram indicating what fashions are being worn by celebrities and social influencers. Fans use the #GetChoosy hashtag to guide the algorithm.

The back end of the design process isn’t the only place where Yoox has implemented AI.

Recently the e-tailer launched a mobile styling suite which allows customers to swipe through different fashions. The customer is guided through selections by an AI named “Daisy” which can try on the fashions for the customer to approximate how they will look.

DISCUSSION QUESTIONS: Does it make sense for apparel brands to use AI to aggregate trend data and use it as a foundation for apparel design? What do you think of the strategy of having a separate, AI-designed line alongside more traditionally concepted collections?

Please practice The RetailWire Golden Rule when submitting your comments.
"It certainly makes sense to integrate trends gleaned from AI into future lines. "
"Trend knowledge does not replace experienced product design and execution."
"When you plan your future artificially, you get an artificial future! Thus endeth the reading."

Join the Discussion!

12 Comments on "Yoox looks seeks AI’s help for design inspiration"

Sort by:   newest | oldest | most voted
Bob Amster

The potential use of AI for the design of apparel is truly a replacement for the real “I” of designers and merchants of recent years. Whether it works or not remains a question. For example is it likely that AI-based designs will appear too sterile and devoid of passion? Is it also likely that some designs may come out looking like a bad joke? The use of AI-based design may be pushing the envelope.

Jennifer McDermott

It certainly makes sense to integrate trends gleaned from AI into future lines and hopefully creating more of what consumers will want from the outset will help decrease the fast fashion landfill problem. As far as differentiating the AI line from the traditionally conceived collections, which are also created using predictive trends (just a more manual version), that feels like a pure PR play.

Art Suriano

There might be a place for this type of AI use however I don’t see it being a huge success. Imagine if we took the same idea and used AI to come up with the next hit song, Broadway show or movie? Fashion design is a business based on new innovative styles, marketing, and hype. Its success comes from designers who have “original” design ideas, and when developed and marketed correctly, those designs appeal to the public until the next big idea. AI has a place, and I don’t mean to dismiss its value, but I fear we’re getting a little too caught up in just how great and useful it is going to be. The human mind will always be the source for new ideas and creativity.

Jeff Sward

There’s an interesting experiment here. An apparel line based on algorithms or a different line based on human design made smarter by enhanced review of the market. I’ll bet on the humans every time. And I don’t think it’s the current trend OR pioneering new trends. It’s AND. It’s the artful, human combination of current AND new.

Ryan Mathews

Using AI for design makes all the sense in the world when we are talking about fast fashion. In theory it allows you to create designs most likely to sell in the moment. Balanced with human designers that can see past the moment, it could be a powerful tool. If however it is being used to predict next season’s fashions — as opposed to addressing contemporary demand — it could be a disaster, since the market will move faster than the algorithms.

Camille P. Schuster, PhD.

The AI approach makes sense as part of the strategy for determining fashion trends. However, even using AI requires decisions about which data to use, what weight is given to what data, and how text comments are incorporated so this is not a machine-only decision. But AI is an incredibly useful tool for this purpose.

Ian Percy

When you plan your future artificially, you get an artificial future! Thus endeth the reading.

Part of me understands that AI is just something we think we should do because we can. Another part of me is saddened that we minimize the miracle of human brain creativity and imagination.

Here’s the brutual reality: ALL data is about yesterday. A “trend” is something that has already happened. What sold isn’t necessarily a predictor of what will sell. In the same way a “problem” belongs only to history and a “possibility” is found only in the future. Net is…AI always looks back. Is that really how to run a fashion line?

A final vent. AI supposedly will “analyze” style trends. What, though, created those styles? It was the amazing unlimited human brain! So Yoox, if your destiny is to play second string…click away.

Doug Garnett

True style breakthroughs’ will only come when an inspired designer sees an opportunity no one else – including the AI everyone else uses – can see. We need to remember that algorithms cannot produce differentiation for the long run. If everyone relies on algorithms then no one with be different.

That said, if a company has a huge line and needs hundreds or thousands of new styles each year and does not need them to be groundbreaking, then I can see filling in the low end of a product line with styles derived from AI.

As to making it a separate line? Consumers don’t care what or who designed it — they want well-made clothes that are attractive and priced as they need them.

Mohamed Amer

With sufficient rounds of “training data,” AI can impress by identifying patterns, optimizing resource and outcomes, and offering predictions for customer behavior based on previous actions, all while processing massive volumes of a variety of data. It’s not (yet) ready to apply – with equal efficacy – those learnings into previously unknown and novel situations.

The more brands work and experiment with AI to improve the apparel design process, the better the outcomes will be over time. Success will come to those who can integrate rather than isolate the power of AI and machine learning algorithms not only into new designs, but on accelerating the design feedback loop so better and faster decisions are made. This will result in greater efficiencies in the supply chain, less waste, higher revenues and potential profits all while exceeding customer expectations.

Jeff Sward

Integrate and accelerate. Perfect.

Anne Howe

AI-aided apparel design is a very cool and efficient way to gather and sort insight and trends. But fashion itself and especially fashionistas are fickle, so there needs to be a healthy dose of human passion woven into the model!

Cynthia Holcomb

Using AI to solve for aggregated trend data is a misnomer for crowdsourcing “trend” data. Listening to fashion commentary across channels still results in style and feature data up for subjective interpretation by apparel brands. Generalized inputs resulting in segmented outputs. This AI use case seems best suited for marketing the current trends, rather than building a well-conceived, branded collection of apparel. Trend knowledge does not replace experienced product design and execution.

"It certainly makes sense to integrate trends gleaned from AI into future lines. "
"Trend knowledge does not replace experienced product design and execution."
"When you plan your future artificially, you get an artificial future! Thus endeth the reading."

Take Our Instant Poll

INSTANT POLL: Do you see AI-analysis of content across social media and online magazines offering value to fashion designers?

View Results

Loading ... Loading ...