In the current marketplace, there are numerous examples of high tech products assembled based on historical market preferences for options (e.g., consumer electronics such as CD/DVD players and TV's) or delivered to customers based on their selected option preferences (e.g., personal computers).  A probable cause for the proliferation of consumer options in configuring products is that customization has the potential to increase customer satisfaction, and may even increase market share through differentiation.  Further, given the short product life cycles characterizing high tech assembly or delivery operations, a firm can keep pace with changing market demands by using alternative options to quickly reconfigure products.  From an operations perspective, however, such strategies place increasing demands on the assembly and logistics processes since customized demand still has to be satisfied quickly.  Firms also need to plan requirements policies to manage the stock of options required for assembly operations.  These policies need to not only address the fact that product demand is often uncertain, but also that market preferences across a spectrum of options may be difficult to accurately predict prior to component manufacture or procurement.

The focus of this paper is on determining the requirements of different component options of a modular high tech product in an uncertain environment. We explicitly model two distinct sources of uncertainty: stochastic product demand and unknown market preferences for the different product options available.  Our cost minimizing model focuses on determining the optimal requirements policies for component options that meet a pre-set service level. We show that simple common-sense requirements policies are not generally optimal; and that there is a non-linear connection between service level and requirements that is hard to characterize without a detailed analysis.  Further, we obtain some counter-intuitive insights into the impact of demand and market preference uncertainty on the optimal requirements.  For instance, we show that an increase in demand variability could lead to smaller component requirements.