The impact of design debugging on new product development speed : the significance of improvisational and trial-and-error learning

Investigating the antecedents of cycle time reduction is a continuing concern within new product development (NPD) research (Chen et al., 2010; Cankurtaran et al., 2013). A number of researchers have reported the effects of team learning on NPD speed (Dayan and Di Benedetto, 2008; Cankurtaran et al., 2013), while others relate learning to overall team performance (Magni et al., 2013). However, few studies have systematically researched the effects of improvisation and trial-and-error learning on NPD cycle time. The aim of this study is to shine new light on NPD learning and cycle time reduction through an examination of the effects of improvisation and trial-and-error. To that end, this study conceptualizes and tests the settings wherein improvisation and trial-and-error might contribute or hinder NPD cycle time reduction. The authors develop hypothesis to investigate the effects of improvisation - and trial-and-error learning on NPD cycle time. Based on a review of the literature and in-depth interviews measures are defined to approximate improvisation and trial-and-error using secondary data from over 200 projects with absolute objective measures of cycle time. In addition, 1000s archival records of debugging incidents and engineering changes are used to approximate the impact of improvisation and trial-and-error. To estimate their impact on cycle time a learning curve model is developed (Argote, 2012) which offers an effective way of identifying the conditions that drive cycle time learning and performance (Wiersma, 2007). Based on this model the hypotheses are tested. The findings suggest that improvisation and trial-and-error contribute to cycle time learning in the prototyping and pilot phases only, and that they hinder learning during later stages in the NPD process. These findings contribute to the extant literature by providing an important new organizational learning perspective on NPD speed. The study contributes to practice by relating firms’ improvisation and trial-and-error practices to learning and speed performance.