Fashion thrives on looking ahead, on novelty. Prediction is her superpower. But how can humans who only see into the future compete with TikTok or the sheer volume of information that artificial intelligence (AI) can process in a short amount of time? Is the trend researcher’s job on the brink? To answer this question, FashionUnited spoke to Avihay Feld, CEO and co-founder of Browzwear, who has been a pioneer of digital technologies since the late 1990s, enabling the fashion industry to thrive in an increasingly digitized world. We also spoke to Tessa Mansfield, Chief Creative Officer at trends and insights experts Stylus.
For years, trend researchers held an unrivaled position in the fashion industry. Brands, from luxury labels to global mass market giants, looked to them for guidance for the future. Trend consulting agencies such as Peclers Paris or Trend Union produced exuberant catalogs of seasonal inspiration, featuring beautiful compositions of color palettes, fabrics and concepts on every page, often punctuated with physical prints, color charts, threads and buttons. These were complemented by photos of stylish people in Tokyo, London and Berlin, taken by the agency’s traveling reporters. Their job was to travel the world with a sizeable budget and track down trends. These volumes sold to brands for several thousand US dollars.
But with the advent of AI, any designer has the power to conjure up any combination of styles, characters, and concepts with just a few taps on the computer screen, while social media provides real-time updates on underground movements, emerging subcultures, and street styles around the world. This speed has meant that trend cycles are becoming ever shorter, as soon as many have spotted something on social media, it’s already a thing of the past.
Feld believes that any brand still trying to predict a trend 15 or 16 weeks ahead is completely out of date. “There used to be predictions, a prophecy that came true and became a trend because a lot of brands followed trend predictions and got the same information,” he says. “It’s going to be the colors, silhouettes, prints, and each brand took something from that, so it wasn’t a surprise to find similar pieces in stores.”
It was sort of a self-fulfilling prophecy. But was that really what the customers wanted? Feld argues that with 60 to 70 percent of the garments made that remain unsold at the end of the season, the answer has to be no. This was evidence of a failed system. Even when these garments go on sale, a large percentage ends up in landfill. “You go out with an order that’s millions of pieces per style, and nobody knows if it’ll work,” Feld said. But the industry stuck to this system. At least until technology called it into question.
One answer, according to Feld, is the model used by digital native brands that “produce for the trend”. This process involves scouring social media, specifically TikTok, and analyzing everything users share and post, taking into account location and demographics, and distilling trends from the data. “Image processing and deep learning are now so advanced that they can understand not only silhouettes and colors, but also the texture of a garment in detail,” he says. Once the designers have evaluated the data and designed their own version, the brand makes a physical sample to see if the style works, and then produces small batches in a short amount of time, often using existing materials to ensure that Less clothing ends up in landfill. “In less than 4 weeks, the garment is in the app, so you can hit a trend in the middle of the season. If it doesn’t work, they drop it,” says Feld. “The risk is low because they only made a few pieces. The ability to produce in line with trends sets them apart from normal fashion brands.”
Is AI supplanting the job of trend researchers?
Mansfield doesn’t deny that a combination of AI tools evaluating social media could provide quick results on fashion trends. “While there is a place for AI-generated, social media-driven product development, it’s all about the ‘here and now’ and is therefore limited to very responsive fast fashion,” she says. These short-term trend-seekers have the potential to cannibalize each other as they compete over the same crumbs. The process of automatically extracting data from TikTok, Instagram and Youtube is commonly known as ‘social scraping’ and it only scratches the surface. According to Mansfield, it is precisely this inability to drill down that gives trend forecasters an edge over AI. “A lot of emerging trends emerging from youth aren’t easily spotted through social scraping,” she says. “You might be able to find the mainstream ideas, but you’re unlikely to find the diverse underground aesthetics and nuanced drivers.”
Technology has democratized every element of the industry, often rendering the ‘gatekeepers’ or those who claim to hold the key to knowledge redundant. But Feld’s and Mansfield’s differing views show that trend forecasting is not a one-size-fits-all solution, but depends on specifics such as brand identity, market position, and customer expectations. Mansfield believes that trend forecasters do not play a secondary role, but rather are an invaluable asset to industries looking to adapt and successfully strategize to compete in these times of tremendous change.
A ‘digital native brand’ has the ability to react immediately to what people are wearing on the street. Feld estimates that such brands update their app with 700 to 1000 parts per day and customers only come to see what’s new. It’s a modern way of browsing during your lunch break or window shopping after work. “It’s the daily fear of missing out,” he says. “But it’s a completely different experience for users than on traditional fashion brand websites, where you get maybe six updates a year.”
Fast fads versus long-lived trends
Mansfield rejects the thesis that all consumers are chasing fast fashion trends, as well as the notion that trend cycles are shortening. “The biggest trends are becoming more and more evolutionary, macroeconomic and longer-term. At Stylus, our clients want to understand how to meet the rapidly changing demands of consumers – across work, active lifestyle, digital and ‘real life’ identity – and most importantly, how to relate to broader contexts such as the cost of living , the imperative of inclusion and the sustainability crisis.”
Feld agrees that some brands set trends while others capitalize on them. And the digital natives he’s referring to, the Sheins and Amazons of the world, haven’t built their reputations on sustainability. They are often in the headlines for producing cheaply produced clothing that does not involve living wages or labor rights for the garment makers, and can also encourage a culture of throwaway clothing due to easy return procedures. The landfill that was avoided in the sample might turn up later.p>
“It’s addicting,” he admits. “But I’m not here to get masses of people to stop eating sugar, I don’t know how to go about it. As long as people are buying, it will continue.” He emphasizes that his message is not that trend forecasting is dead, but that it needs to evolve: “You don’t need 15 weeks to set a trend. You may not be able to do it in 4 weeks, but the truth is in between.”
Mansfield’s clientele, some of the biggest global brands, have different expectations, and the advice she offers goes beyond product development and even innovation to include a look at the big picture and at long-lived trends. She sees the role of trend forecasting as connecting the dots between emerging trends, looking at the full range of consumer lifestyles and highlighting opportunities for cross-industry collaboration and inspiration.
“The forward-thinking brands and agencies we work with understand that trends don’t exist in isolation,” she says. “Most fashion brands still have to plan a year and a half to two and a half years ahead – especially in categories like Active – and that timeframe can’t be determined by scouring social media for inspiration. Inspirations that can come and go in literally a matter of weeks just tell you what’s already happening.”
This translated and edited post previously appeared on FashionUnited.com.