Skip to content ↓

Research collaboration and math model guided intern through 'fast fashion' world

Juan Correa spent six months as an intern at Zara, an international Spanish clothing manufacturer. But don't ask him if short skirts are "in" next season--the graduate student in the Leaders for Manufacturing Program (LFM) wouldn't know.

Instead, Correa, a native of Mexico City, devoted his LFM internship to implementing a sophisticated optimization model using a mixed integer mathematical program at Zara's headquarters in La Coruña, Spain.

Zara is part of Inditex, one of the world's largest fashion distributors. Zara, which has eight distribution lines and more than 900 stores worldwide, features inexpensive fashion for women and men mainly between the ages of 16 and 35. Its chief competitor is H&M, the Sweden-based store that also sells trendy clothing for low prices. Zara has 24 locations in the United States.

The Inditex business model is notable for its vertical integration and short turnaround time. Thanks to this "fast fashion" model, Inditex clothing can leap from drawing board to retail rack in as little as six weeks. Another retailer might spend six months on the same process, according to Correa.

Correa's internship at Zara was a business model in its own right. It resulted from a research collaboration on inventory optimization that included Correa and Jérémie Gallien, associate professor of operations management at MIT Sloan, on campus, and two MIT Sloan alumni, -Felipe Caro (Ph.D. 2005), now at UCLA, and Inditex special projects director José Antonio Ramos, a 2001 MIT Sloan graduate. Gallien and Caro began their collaboration with Zara in August 2005.

Through 'fast fashion' to global agility

According to Gallien, Zara's "fast fashion" has inspired several ongoing and previous research projects at MIT Sloan. Zara is one of just a "few global retail brands with the operational ability to change their store assortment on a weekly basis, as opposed to only once per season, as many U.S.-dominant retailers do," he noted.

Zara started where those stores did. Its inventory distribution process, established when Zara had only a few stores, was overwhelmed once it had nearly 1,000 stores. Correa took on the challenge of building an updated, -forward-looking and agile distribution process.

To illustrate the shortcomings of the "few stores" model, Correa explained that every week, each Zara store would place its order at one of the two warehouses in Spain, with no limit to quantity. The result was that some stores were not getting enough, while other stores received too much. Unsold clothing piled up in some locations, while inventory was rapidly depleted in others. The company needed a better distribution system.

Gallien and Caro developed an optimization model, which Correa had to implement. "They handed me the program on a piece of paper and said, 'Let's make it happen,'" Correa said.

At MIT, "make it happen" is not a theory--someone has to live with the problem. Correa was that someone. He saw the old system's problems firsthand, as he spent the first two weeks of his internship shadowing a store manager in the men's department at a Zara store in Dallas.

He discovered the uneven inventory numbers among stores resulted from the manual ordering process run by store managers. The managers submitted store orders via computer, and since there was no limit, they received as much inventory as they requested. When there wasn't enough inventory available, warehouse staff had to calculate manually which store was to receive what supplies. Unfortunately, the individual store managers often could not predict sales accurately and had no information on what the other stores needed, Correa said.

He gave another concrete example of the problem: When 300 Zara stores ordered a total of 2,000 pairs of pants and the warehouse only had 1,500 pairs to distribute, there was no reliable process for allocating the pants.

Correa's project was creating a forecast, based on historical data, to devise a demand/estimate figure for each store, so sales across the board would be maximized. Each warehouse was then directed to ship inventory based on the calculated demand figure devised by -Correa.

The concept of ordering the goods based on analytical history and mathematical concepts was entirely new to Zara. "In the fashion industry, everything is based on intuition. You can't put numbers behind it. Instead, it's, 'I think this skirt will sell.' This is one of the things the owner of Inditex does extremely well. He has an amazing sense for what sells and what doesn't. But we had to put some numbers and quantitative analysis behind those decisions of where to ship the merchandise," he said.

Correa, who was a process engineer at Texas Instruments before coming to MIT Sloan, had had no experience in the garment business, which he characterized as "quite a long way from the semiconductor industry."

Zara was also a long way from the cubicle environment, Correa notes. At Zara headquarters, everyone sat in one large room, grouped together at open desks. The outer edges of the room--where the designers sat--were ringed with clothing racks formed according to section--women, men and children. Country managers sat in the middle, and the executive management team shared one large table in the middle. When designers debuted outfits for a new season, they held a fashion show for the entire office.

Correa's internship at Zara was a cross-cultural experience as well, he said. It took him a little time to adjust to the company meetings--instead of speaking in turn, everyone in the meeting virtually spoke at once. "At first it's a little overwhelming," said Correa, who speaks fluent Spanish. "But it's an open environment, so everybody's opinion matters. The organization is pretty flat."

A version of this article appeared in MIT Tech Talk on December 20, 2006 (download PDF).

Related Links

Related Topics

More MIT News

Headshot of Catherine Wolfram

A delicate dance

Professor of applied economics Catherine Wolfram balances global energy demands and the pressing need for decarbonization.

Read full story