Management of the supply chain in a rapid product development environment

Ethicon Endo-Surgery (EES), a Johnson & Johnson company, has experienced consistent supply problems when introducing new products. On several occasions, demand for a new endoscopic surgical instrument has outstripped EES's ability to supply the insbllment. (Demand in this sense refers to the volume of product ordered by distributors.) Depleted finished goods inventories have translated into several mOliths of continuous stockouts. In response, production has been forced to quickly ramp up its output in an effort to restore consistent supply to distributors and, consequently, to end·cu~tomers (hospitals). Ethicon Endo-Surgery's strategy of rapidly developing innovative products and of servicing hospitals through new sales channels has earned it tremendous success in the market for endoscopic surgical instruments. In keeping with EES's product development strategy, the original research problem focused on forecasting demand for new products. Having consistently experienced supply shortages for new products, EES felt that being able to better predict distributor demand could help alleviate the supply problems. In other words, EES wanted to understand the extrinsic factors associated with new product introductions. Interestingly, the extrinsic factors did not tum out to be nearly as important as the intrinsic factors. Distributor demand was found to be surprisingly predictable and stable relative to production. Variations in production rates from week to week and from month to month often exceeded the corresponding variations in incoming demand. And despite the relatively stable demand, finished goods inventory levels were found to be well in excess of corporate objectives (on the order of $10 million). The real issue, therefore, was not in predicting distributor ordering patterns, but in the way that EES convened the orders into actual production plans and inventory policies. In this thesis, I provide the research path that began with an extrinsic view of new product introductions and ended with a focus on processes within EES. I will demonstrate that distributor demand patterns are relatively predictable, even during new product introductions. I will then show how month-to-month changes in production exceeded month-to-month changes in demand for both new products and stable products. An analysis of finished goods inventory levels uncovered inventory well in excess of corporate