Order quantities for style goods with two order opportunities and Bayesian updating of demand. Part I: no capacity constraints

This paper is the first of two that study the problem of ordering a family of style-goods products where demand is uncertain and there are two order opportunities. The first opportunity has a long lead time and low unit cost. The second opportunity has a short lead time and high unit cost. During the time between the two order opportunities new information on demand becomes available. The information is used in a Bayesian estimation process to revise demand forecasts. There are no capacity constraints at the order opportunities. The second paper extends the results in this paper to the situation where there are capacity constraints. A number of inventory models having different information and computation requirements can be used to determine good order quantities. We find that complex models are appropriate for the most important A items. Simple models are best for other A items and for B and C items. The motivation for studying this problem is the experience of a real company. PTK has one medium-size factory and a chain of retail stores in Canada and the United States.