fashion technology

State of play: Trend forecasting in the fashion industry

Key takeaways

  • The present ways of fashion trend forecasting rely unanimously on runway collections and on the past to predict the future.
  • The modern fashion market has changed: time-to-market is faster, knowledge of foreign markets is essential, and design functions and analytical departments have no choice but to be aligned.
  • The combination of image recognition, forecast algorithms, and machine learning comprise what is today’s ultimate trend prediction solution.

To be ahead of a trend is to have the upper hand, so brands who forecast the market well enough to stay two steps ahead are best-positioned for success. But without a crystal ball, how is this possible? How can fashion brands see into the future of the market?

Trend forecasting today: Room for improvement

Historically, fashion forecasters attended runway shows during fashion weeks to take note of designers’ collections. These forecasters then provided department stores and major chains with their trend predictions, and these stores would thus base their collections off of what debuted at Fashion Week. Twelve months later, the general public would be sporting whichever trends trickled down from the year prior.

But today, trend forecasting no longer operates at the mercy of designers’ collections. There now exist new ways of predicting the season’s top colors, patterns, and shapes. Fashion and luxury houses hence rely on certain tools to drive their collections, namely:

  • Past sales analysis
  • Feedback from stores
  • Archives from their own collections as well as those of their competitors  
  • Industry trade shows 

Looking to store feedback and past sales, for example, is no longer fast enough to keep up with trend propagation and product perception at the hands of influencers. This data is essential for brands still today, but the market now moves too quickly and is too democratized to wait for trends to become obvious. Social media platforms like Instagram and Weibo are merciless in their turnover-time, with some trends lasting 2 weeks before going out of style.

Heuritech Podcast Series

Why stay up-to-date, when you can stay ahead

Before a brand can explore new trend and demand forecasting solutions, there are several realities to understand:

Time-to-market speed is increasingly important. Reactivity is sometimes more important than being accurately on-trend, because so many brands churn out new pieces and collections at a rate the fashion industry has never before seen. Brands must thus understand how to transform their existing supply-chain model into one that fits the modern, internet-driven market.

Knowledge of foreign markets is no longer optional. Today, understanding foreign markets is vital to expanding one’s brand. Among others, the Chinese market promises the most potential, with luxury brands already putting up stores and collaborating with local designers across the country.

Design functions and analytical departments must be aligned. Design processes, merchandising strategies and buying planning are interconnected. All three functions must confirm each other’s intuitions to achieve greater commercial success.

The present ways of trend forecasting rely unanimously on the past to predict the future. So what if brands could rely on the present to predict the future? 

Read more

Report | SS ’25 Men’s Fashion Week

Download the report

Trend forecasting solutions for fashion brands

Every day, people around the world share 100 million images and 400 million stories on Instagram. The power of these posts is undeniable, as influencers and non-influencers alike provide visibility to brands when they post their newest bag or shoe. Technology that is able to leverage this content thus seems like the logical next step. 

Other companies have caught on, but often their solutions fall short. 

While Heuritech is not the only company to use Machine Learning, Image Recognition, and forecast algorithms, it is the only company that does so comprehensively.

Current solutions often analyze only text and hashtags in social media posts, leaving out a vital aspect of social media: shared images in which users showcase their newest outfits, often with no textual mention describing what they are wearing.
Otherwise, they often analyze images from catalogs published online, typically on the brands’ website or major fashion news sources. These images do not reflect customer reception or perception of the products.

Heuritech goes beyond these current approaches.

How? By applying advanced artificial intelligence to translate images and videos shared into meaningful market insights. From Instagram alone, our technology analyzes over 3 million images daily combined with the ability to identify more than 2000 attributes, within each image,  regarding patterns, shapes, fabrics and colors.

Thanks to this great volume of data, combined with historical analysis, geographical breakdown and client segmentation, Heuritech is able to predict and quantify each trend behavior up to 12 months ahead. 

Heuritech graphic of client benefits

In the age of social media, the traditional business model is no longer sufficient. Brands must now interact with their customers, welcoming them into their world via social media. Consumers are now able to be publicly vocal about what they do and don’t like, with shared images and videos providing brands with tangible feedback on their products. Analyzing the goings-on of social media is thus vital for the success of a brand, but without help, this task is Herculean. Fortunately for the fashion industry, a trend and demand forecasting solution is available.

photo profil de l'auteur, Mélanie Mollard

About the writer: Mélanie Mollard, Fashion Content Writer

Mélanie writes about the fashion industry and its many features through the lens of AI and applied data.

Questions or feedback? Email us at

Get contacted by one of our experts

Blending artificial intelligence with fashion expertise for predictive analytics on trends

Press inquiries:


© 2024 ALL RIGHTS RESERVED - HEURITECH | PRIVACY POLICY | TERMS & CONDITIONS | 66 avenue des Champs-Élysées, 75008 Paris, France