When you hear the term “big data,” the first thought that comes to mind likely isn’t the world of high fashion. Yet data science is reshaping fashion merchandising strategy, in some cases well beyond its effects on traditional data-driven industries such as finance, manufacturing, and medicine.
Swedish retailer H&M, the world’s largest clothing brand, recently revealed plans to factor big data into sales forecasts, believing that improved predictions will help the company stock merchandise more efficiently across thousands of stores. H&M is just one of the fashion companies investing in smarter tools to capture and apply vast quantities of data to enterprise forecasting and planning.
Big Data at H&M
The rise of online shopping and the advancement of fiercely competitive digital startups have been eating away at H&M’s same-store sales for more than two years. Amply aware that it must charge life into it’s brick and mortar business, the company has made plans to capture and analyze data from a variety of newly usable sources including store receipts, return item records, and loyalty-card usage. Each of these data streams will help better gauge customer demand, management believes.
At one H&M store in Stockholm, for example, management has recently revamped its offerings to better serve its primary clientele. After discovering that most of that particular store’s customers were women, the store cut back on the amount of menswear it offered and stocked on popular women’s items such as floral skirts in pastel colored apparel. The company says that sales at this store have “improved significantly” as a result of the changes. This is in direct contrast to H&M’s traditional approach to merchandising, which was to stock all locations with roughly the same fare, regardless of substantial differences in location-based demographics and geographic variables.
Not only does H&M analyze customer behavior at its individual stores, it also uses artificial intelligence and machine-learning technologies to scan for the next hottest fashion trends. By analyzing millions of posts on blogs and social media using computer vision and natural language processing, H&M data scientists can better understand what their audience is purchasing, and in which types of products they’re interested.
Big data across the fashion industry
With this latest announcement, H&M joins dozens of other large fashion and retail companies looking to capitalize on the reams of data that they’re generating from customer interactions.
Big data startup Editd, for example, consults with fashion companies that are looking to perform many of the same analyses as H&M. Editd, which counts Target and Gap among its clients, has aggregated more than 53 billion data points from a variety of sources going back more than four years. Data sources include social media, runway reports, and a tracking list of more than 1,000 retailers around the world. Another Editd client, the clothes retailer Asos, found that its sales increased by 33 percent after working with the market-data consultancy.
The potential applications for big data within the fashion industry are numerous:
- Identifying cross-selling opportunities by discovering which items are frequently purchased together (such as the same purses and shoes)
- Recognizing the factors that make people more likely to return an item, and placing limits on returns for customers who attempt to abuse the system
- Analyzing your sales pipeline in order to understand why some potential customers fail to convert right before the purchase
- Finding promising newcomers to the fashion industry just before they strike it big so that you can collaborate with them and feature their items in your stores
Ultimately, leveraging the power of big data is all about better understanding and predicting human behavior. This makes it the perfect fit for domains like the fashion industry, which is always looking ahead to find the hottest, newest insights and trends.