Continuous Improvement Training Program
A Continuous Improvement Program
Featuring 8-Training Modules by Bill Gaw

Increase the effectiveness of your
Lean Manufacturing Training Program
Lean Manufacturing Simulation Game 

Case Examples

Here are two case examples of forecasting improvement efforts, how they were set in motion, experiences along the way, and the principal outcomes. [More examples and added details will be presented in the conference session.]

A. Catalog Forecasting

J. Sullivan3 was a large mail-order catalog company, em­phasizing apparel, footwear, and similar consumer prod­ucts. Like many direct marketers, growth in its business had slowed in the 1990s after a decade of rapid industry growth. The market had become saturated, with many consumers receiving one catalog per day from various mailers. In their efforts to attract customers, J. Sullivan and others had introduced new catalogs, varied the sizes of catalogs, changed layouts, added new products, and made many other changes. One result was poorer forecasts. With so many aspects of the business changing from one catalog to the next, it became harder to predict sales. In addition, much time was wasted in management commit­tee meetings, as members debated the pros and cons of different ways to improve response rates.

In this case, a straightforward regression model was able to sort out and quantify the contributions of many market­ing factors that affected sales. Examples included the mix of people who received the catalog, format and size of the catalog, timing of the mailing, and the products offered. In one single catalog, the model allowed management to save $400,000 in production and mailing costs.

B. Adapting to the Longer Lead Times of Far Eastern Suppliers

Sport Obermeyer was a Colorado-based producer of skiwear. In a highly seasonal business, and with a strong fashion component to its sales, it found itself in the unhappy situation of facing longer lead times as a result of its movement toward Far Eastern suppliers.

Management took several steps to improve its forecasting. First, it arranged its own private "show" for retailers, about 6 weeks before the traditional ski industry trade show where most producers introduced their new styles for the coming year. Second, Sport Obermeyer phased its produc­tion schedules according to the uncertainty of demand for products. Items whose sales could be predicted with reasonable accuracy were scheduled for production far in advance of the selling season. Manufacturing capacity was thus kept free for responding to the more uncertain items after the private "show." The company developed a com­puter model to develop forecasts based on the show results and to optimize the timing of production in relation to expected sales.

Conclusion

When customers demand faster response, their suppliers have three basic choices. They can carry higher invento­ries, a strategy which American industry has generally not pursued. They can redesign manufacturing and operations processes, the principal choice in recent years. Or, they can improve their ability to anticipate customers' needs. For a variety of reasons, most comp anies have not exploited their opportunities in this latter option. As this paper has illustrated, forecasting can be improved, at reasonable costs and with important benefits in better understanding of your business. In fact, given the payoffs, many compa­nies that already have moved to JIT, flexible manufactur­ing, or Quick Response can further improve their service levels and costs by taking a careful look at their forecasting function.