How Does Social Media Shape Fashion Trends?
Predicting a fashionable future, one hashtag at a time.
Predicting a fashionable future, one hashtag at a time.
In our digital age, the fashion industry is moving faster than ever, and brands need to react with a quicker time-to-market to not only survive, but to thrive. In this context, social media is undoubtedly the luxury industry’s pot of gold waiting at the end of the rainbow. But how does social media shape fashion trends? And moreover, how do innovations in social media affect the fashion industry?
Needless to say, Instagram has been a game-changer for the fashion industry since its creation in 2010 and today brands are relying on its use to predict future trends and market their own products into trendy must-haves. Paris-based tech agency Heuritech has de facto used Instagram to develop its own digital tool for trend-forecasting – one that predicts the behaviour of trends and the performance of key luxury products. The start-up company de facto conceived a digital image recognition tool, which predicts what fashion product trends are coming and how they will perform based on pictures sourced on Instagram. The recognition tool analyses a myriad of aspects, such as colours, styles, textures, patterns, shapes, and logos. In this context, over three million pictures are analysed each day by the AI-powered tool, resulting in over 4000 identified trends, 2000 identified fashion details in any image, and a product recognition efficiency with a 90% precision rate. “Today, more than ¾ of purchases are influenced by this social network,” stated Tony Pinville, CEO and Co-Founder of Heuritech, adding that his trend-forecasting company also focuses on Chinese social media channel Weibo, and is currently expanding into TikTok, which is highly popular amongst Gen Zers. “Our data scientists’ mission is to identify early trend signals on social media from the very beginning to forecast what is going to happen. We want to push this analysis further with brands willing to collaborate with us on their sales data,” Pinville further explained.
This raises the question: is this new approach to trend-forecasting on social media about to disrupt our fashion industry? In fact, it could. AI-powered image recognition tools do not only allow data scientists to discover trends on social media at an early stage and observe their evolution, but are also used to support brands in their decision-making, notably in optimizing their product assortments, forecasting their product quantities, and mastering their product positioning and marketing strategies. As of today, a handful of stablished international trend-forecasters dedicated to the luxury, fashion, and lifestyle sectors, such as the aforementioned Heuritech, but also WGSN, amongst others, are setting their sights on social media-based trend-forecasting. “At WGSN, we have proprietary tools that help us analyse the social media landscape, but, in the future, with developments on AI and visual recognition, more seamless way of reading the patterns within imagery will be the norm,” explained Sarah Owen, Senior Insight Editor at WGSN. Owen further stated that social media has fundamentally changed trend forecasting in regards to the establishment, speed and distribution of trends. “It’s now a much more cyclical process where online and offline inputs simultaneously create a new lifecycle of trends,” she added. “Social media, specifically visual platforms like Instagram and TikTok, have become research touchpoints that we reference often.”
The question arises as to which social media accounts are considered as research touchpoints by trend-forecasters? Amongst countless social media accounts dedicated to fashion and luxury, only a dedicated selection of clusters meeting specific criteria of relevance and quality are taken into account when analysing trends on social media. “By capturing early signals from our ‘edgy’ audience panel on Instagram, we can forecast expected growth from the fashion-forward to the mass,” explained Célia Poncelin, Heuritch’s Head of Marketing & Communications. The so-called “edgy” audience panel refers to Instagram content shared by the top-performing professionals from the luxury industry, mostly cherry-picked from the BoF 500 list, but not only. “The goal is really to be representative and neutral about the people who make today’s fashion industry. We do manual curating with fashion experts but also automatic curating, by looking at five years of historical data,” she added.
However, there are some limits to this “picture-perfect” Instagram-based forecasting and its ultimate relevance when predicting future industry trends. “We’re living in a post-trust society where consumers are sceptical about the information they’re seeing and reading. As a result, disclosing paid partnerships and posts is critical in building integrity within communities,” stated Sarah Owen. But whether posts are labelled as advertising (#ad) or not, is often left to the influencer’s personal choice and morals: in fact, paid product placements are not always labelled as “sponsored posts” on Instagram. These hidden ads de facto constitute a challenge for trend-forecasters, who need to make sure that the social media data used in their analysis is based on a post’s performance that is organic and not sponsored — otherwise, the results cannot be trusted. A challenge that most trend-forecasters still grapple with. This also implies that fashion brands need to focus on the adoption of their product by real consumers, and not only paid influencers, to guarantee that their product is trending in real life. “When a product’s visibility only comes from big influencers, and is likely to be inflated by paid product placements, there is no quick adoption from micro influencers, which means that there is a very high probability that this product will not be a commercial success and that its hype will quickly fade away,” Pinville concluded. The luxury industry’s next battle horse is one of transparency – on social media and beyond.