In the fashion industry, Artificial Intelligence (AI) not only enhances the accuracy of clothing recommendations, but can also generates critical insights into trends, streamlining the design process and enabling precise demand forecasting for hyper-localized products.
In the past there were already experiments in using AI to analyze trends. In 2018, IBM collaborated with Tommy Hilfiger and the Fashion Institute of Technology (FIT) Infor Design and Tech Lab on a project called Reimagine Retail. This initiative aimed to equip future retail leaders with AI-driven design skills and provide inspiration for upcoming trends and forecasts.
FIT students were granted access to IBM Research's AI capabilities, including computer vision, natural language understanding, and deep learning techniques, all specifically tailored for fashion data. The AI tools analyzed extensive datasets, such as 15,000 of Tommy Hilfiger's product images, 600,000 publicly available runway images, and nearly 100,000 fabric patterns. This trend analysis allowed students to develop designs that not only aligned with Tommy Hilfiger's brand identity but also reflective of current and emerging fashion trends.
While the initial collaboration between IBM and FIT focused on leveraging AI to identify trends for new designs, nowadays the broader application of AI in trend prediction is proving invaluable for fashion companies. According to companies using AI for this purpose, it facilitates faster product launches and minimizes lost sales.
WGSN is among the companies leveraging AI to identify trends: the consumer trend forecaster launched this week a new data-driven platform called WGSN Fashion Buying.
This platform builds on WGSN's trend expertise and integrates its proprietary TrendCurve AI predictive analytics. By combining retail data with insights from social media and runway shows, it offers a more comprehensive approach than other retail analytics platforms, which often rely solely on historical data from a single source.
WGSN Fashion Buying is designed to support buyers by providing insights into product direction and emerging trends across the entire development cycle, from pre-planning and development to in-season analysis. The platform promises "future-proof trend decision intelligence" with buying-specific forecasts tailored to each stage of the product life cycle and customized by category.
At the moment, WGSN Fashion Buying covers key categories such as dresses, cut and sew, woven tops, denim, outerwear, trousers, skirts, accessories, and footwear. The platform introduces four new forecasts: Trend Narratives, which identifies key items and colors for building seasonal stories and concepts; TrendCurve AI Color, offering data-backed color assessments and two-year color projections; TrendCurve AI Materials & Details, which tracks details and their usage across various retail segments and TikTok Trading.
The latter pinpoint opportunities to re-merchandise existing inventory, capitalizing on emerging and evolving TikTok trends that align with current seasonal sales.
Essentially, WGSN aids buyers in distinguishing between TikTok's fleeting viral moments and genuine trends that hold commercial potential. It also helps buyers checking out existing products in inventory that might already align with these trends, eliminating the need to add new products. By doing so, WGSN Fashion Buying contributes to addressing the fashion industry's waste issue. Recent research by WGSN and OC&C Strategy Consultants indicates that shifting to a demand-driven planning and buying model could reduce overproduction by 5-15 percent, mitigating a major source of unnecessary waste.
Opinions on AI's role in reducing waste are mixed, though. In May, during the Global Fashion Summit in Copenhagen, at the "Ending Oversupply" session, Dr. Ahmed Zaidi, CEO & Co-Founder of Hyran Technologies, criticized the use of AI to stop overproduction and overconsumption, comparing it to "attaching a jet engine to a broken process".
Besides, there may be another issue with AI-powered trend forecasting tools: while WSGN Fashion Buying platform promises to enhance seasonal planning and reduce decision-making risks, questions remain about its data collection methods.
AI forecasting tools typically rely on extensive datasets to make accurate predictions, which can include scraping images from the internet. This raises potential ethical and legal issues such as copyright infringement and data privacy. To generate timely trend forecasts, WGSN Fashion Buying requires diverse data sources, including social media, e-commerce sites, and runway images, besides, it will also have to constantly monitor and process the contents of TikTok.
However, whether WGSN uses direct scraping or other methods like licensed data feeds, APIs, or partnerships with content creators would depend on their specific approach to data collection, as well as their adherence to legal and ethical standards.
Comments