At the end of a previous post entitled Trading Companies And Efficiency, I explained how some large intermediaries could accelerate the manufacturing cycle, and this way minimize the risks of poor sales forecasts. Actually importers can also do it on their own. It reminded me of an HBR case study that I read, back in business school. Its title is Making Supply Meet Demand In An Uncertain World.

What is it about? It describes how Sport Obermeyer, an American company producing fashion ski apparel, improved its sales forecasts and made its supply chain more easily adaptable.

I am going to summarize the main points of that case study:

The problem: Importers and retailers have great difficulty forecasting in-store demand of fashion items. Poor forecasting can result in either lost sales–if the product is quite successful–or unsold inventory.

How to improve forecasting: the company was relying on consensus-based meetings to predict a new style’s success, with poor results. This process was redesigned as described below:

• Some customers were invited for a preview of the new season’s lines, and early orders were solicited (representing 20% of the season’s volume).
• These early orders were found to be highly correlated with the following orders of the season. So they are a very good indication of the commercial success of a given style.

How to make the supply chain more flexible: the company took several initiatives:

• Forcing their designers and their merchandisers to work together and standardize the materials to purchase (e.g. reducing the variety of zippers used by a factor of five, and buying them all in black color).
• Anticipating on the raw materials to use, buying them in advance, and placing them in warehouses.
• Keeping the materials (before production) as undifferentiated as possible until they have to be dyed, cut and sewn.
• Booking the factories’ production capacity in advance, and them deciding on the exact colors and styles to produce as soon as orders are received.
• Making the most predictable styles first, and saving the capacity for highly unpredictable products at the last moment.
• When the quantity of a style represents many times the minimum required order, it is cut in several production runs that are processed at different times.
• Sending part of the production by plane.

This case study was written in 1994. The company was already buying from China. The standard lead time is about 3 months for development and production, and 1 month for sea freight. With a peak annual season lasting only 2 months, the company had to adopt new methods to respond faster and more accurately to demand. From my observations, most garment importers are still a long way behind in their development and purchasing organization.

Overall, a lot of money is wasted in poor process optimization. There is actually still a place on the market for well-organized Western production of apparel!