Vareska Van De Vrande

New product development (NPD) is a risky and expensive endeavor. Therefore, discontinuing less promising NPD projects, based on their development trajectories, is very important, yet firms face challenges when deciding whether and when to discontinue NPD projects. This article examines how characteristics of a firm's prior success and failure experiences, such as the number of launched and discontinued past projects, their attributes (i.e., the codification and quality of the knowledge underlying these past projects), and their level of relatedness (i.e., similarity) to current projects, influence the risks of continuing and discontinuing NPD projects, thus impacting their future discontinuation decisions. We adopt a mixed method approach, where we formally model and empirically test the influence of prior experiences on future decision‐making. A formal model enables us to offer reasoning that considers the risks of continuing a project versus discontinuing that project as the basis for theoretical arguments for a set of proposed testable relationships. We then empirically test these relationships on 2938 new drug development projects of biopharma firms worldwide. We find that firms with a greater amount of success and failure experiences tend to discontinue less promising NPD projects sooner than firms with less success and failure experiences. Also, we find that while the attributes (particularly knowledge codification) of successful experiences predominantly helped firms decide on timely discontinuation, failure experiences influenced firms' decision on timely discontinuation only when the projects were closely related to their failure experiences.