CAMBRIDGE, Mass., Oct 3, 2007 — The use of mathematical modeling and sophisticated optimization techniques is reaping financial gains for companies and a better shopping experience for customers in the once-staid fashion retail industry, according to MIT Sloan School of Management Professor Jérémie Gallien. Zara, a leading Spanish clothing chain, recently adopted the analytic approach to inventory distribution developed by Gallien and his collaborators UCLA Anderson School of Management Professor Felipe Caro, Juan Correa, and Jose Antonio Ramos, all graduates of the Sloan School.
“The same kind of ‘rocket science’ revolution that occurred on Wall Street more than a decade ago is now hitting retail,” says Gallien. “Because of the volatile demand trends they face, large fashion retail companies were, until recently, driven more by a culture of human judgment, intuition and other ‘left brain’ activities. Now many are also implementing sophisticated optimization models to help them make more intelligent decisions in such important areas as distribution and pricing.”
Zara, which is owned by Inditex, has long relied on the proven ability of its founder and top designers to sense and anticipate market trends, creating an overall decision-making culture that even more driven by intuition than in other companies, making its implementation of mathematical models especially significant. The company is able to get clothing from concept through production and into its stores in as little as six weeks, compared to the six months or even longer it takes traditional retailers.
In September, Inditex beat analysts' forecasts when it reported a 7 percent increase in “like-for-like” sales over the first half of 2007, despite fears of slower economic growth in Spain. A significant portion of that increase could be attributable to the model developed by Gallien, Caro, Correa, and Ramos which Zara deployed that semester. “Performing a controlled pilot field experiment in late 2006 involving a subset of references and half of Zara's store network, we estimated that our inventory optimization model increased sales by 3 to 4 percent,” says Gallien. Zara Chief Financial Officer Miguel D’az confirms: “Our estimate of the financial benefits specifically attributable to the use of that new process when distributing inventory to all our stores worldwide is consistent with that reported for the live pilot.” Díaz also said that this mathematical model “significantly contributes to Zara's strategic goal of improving the scalability of its operations in order to support its continued growth.”
Customers also benefit from the combination of fast fashion and analytic modeling. “It makes the shopping experience much more interesting and rewarding for the customer,” says Gallien. For example, under his model, inventory is sent to stores where it is most likely to appeal to customers, including more complete sets of sizes to spare customers the frustrating experience of finding a garment they like, only to find it unavailable in their size.
Under the “fast fashion” model, product life cycles are much shorter, meaning that customers see a constant turnover of a store's clothing assortment. “If you go to a store such as the Gap, you're likely to see the same items on display two weeks later, if not for the entire season,” says Gallien. “But at Zara, if you go two weeks apart, you'll have a much more interesting inventory selection. On the other hand, if you don't buy something right away, you may not see it in two weeks, because the stores don't stock too much of any item.”
For this retail strategy to be successful, a retailer must align its total operation, including design, production, customer feedback, and distribution,” says Gallien, which could be a challenge for U.S. retailers. Because European retailers such as Zara and H&M are closer to their production sources, they can achieve such alignment more easily. Nonetheless, he says U.S. retailers could create their own fast-fashion brands, possibly turning to Mexico for production rather than places like China or Bangladesh in order to speed up their design-to-shelf leadtime. Provided U.S. consumers would be ready to pay for the fresher product assortments that fast-fashion offers, which remains to be seen, the higher costs of labor could be offset by greater sales due to more frequent store visits by customers and reduced markdowns.
“Zara's successful implementation of mathematical models may still signal an important opportunity for U.S. retailers, because these quantitative methods may be directly applicable to them” says Gallien. “I'd be very surprised if strategic planners and other leaders of major U.S. retail companies were not very closely scrutinizing that company's success.”