WALK FORWARD: Tagwalk is adding trend prediction to its services.

The Paris-based fashion search engine is launching a new AI-powered dashboard that combines the detailed tagging derived from 410,000 images of more than 1,300 brands with insights gathered from the search histories of its 300,000-strong community of fashion professionals.

“We realized we had so much information [in Tagwalk] that there was definitely a way to be able to predict and understand what products were being made and what was performing,” said founder and chief executive officer Alexandra Van Houtte.

Eighteen months in development in-house, the new service takes the shape of a prediction dashboard divided into categories that range from colors and materials to themes, styling and even beauty. Trends are identified thanks to views, performances and rankings that are calculated without human intervention.

“Our audience really managed to give us a new model for prediction and trend analysis that doesn’t exist on the market yet,” Van Houtte said.

Users can also drill down on what keywords and searches have led to specific brands, as well as compare terms and trend behaviors. It also offers a read on the demographics and territories of the users who have searched for these trends.

Tagwalk described the new feature as a solution to combat the “excessive research burden” that led to missed opportunities for fashion businesses and could help tackle overproduction of unwanted items. 

For brands like Adidas, which is a longstanding client of Tagwalk’s existing reports service, the real-time analysis and quality of the data are a major selling point, according to Therese Hermann, manager of hype insights at the sports giant.

“Knowing what our consumers are buying in other places helps us not only offer our [wholesalers] exactly what they need and will be selling, but also strengthen our product teams in their decision making and sell-in process,” Hermann told WWD in an email. 

Van Houtte stressed that the new feature would not replace designers, instead helping brands to “calibrate a little better,” she said.

In addition to designers using the information to orient their creations, Van Houtte pointed out the information could be used by merchandisers to curate a product assortment, but also by social media managers seeking to adapt their upcoming posts to what consumers are looking at.

For the seven-year-old French company, the dashboard’s addition is part of a wider reshaping of its service offering.

While brand searches and look book viewing remain free, keyword searches and mood-board creation are behind a paywall introduced in the fall that’s priced at 100 euros a year, with a deep discount for students. Access to the trend prediction dashboard starts at 2,100 euros a month.

“What we didn’t want is for people to [only] depend on us for consulting [services] and reports, we want people to use us every day as a fully automated SaaS service,” Van Houtte said. “Obviously that’s a game changer for us in terms of [business] but also in terms of tech, as we had all this data that wasn’t really being used correctly or efficiently.”