Sequential Decision-Making in Product Development

Presenter

Jiawei Li

PhD candidate

Technology & Operations

Ross School of Business

University of Michigan

Time and location

1:00-2:00 pm on BlueJeans: https://bluejeans.com/999383152

Abstract

In the automotive industry, the project for the launch of a vehicle usually begins several years before the planned launch time. During the process, each car part (such as seat) will subject to a series of design change opportunities due to marketing, engineering, or safety needs. Such opportunities happen in sequential order, and a decision of whether to conduct the design change or not must be made as soon as the opportunity arrives. i.e., it is not possible to wait and “bundle” these decisions. All of this is managed under a pre-specified budget.

Using laboratory experiments, we implement the above process as a dynamic resource allocation problem. We find that subjects fail to perform optimally and exhibit a large degree of individual heterogeneity. We find suggestive evidence that this is due to subjects’ failure to jointly consider earning benefits and spending budget. Three diagnostic tasks are introduced to help us capture the aspects of individual heterogeneity that are relevant to this problem: risk preference task, CRT, and Hit 15 - a task intended to test people’s ability to conduct backward induction.

More info on joining through BlueJeans