More and more companies are actively involving their customers in the new product development (NPD) process. However, there is little consensus regarding the contribution of customer involvement to new product outcomes. A better understanding of this contribution can shed light on whether and when it is worthwhile to involve customers and thus provide firms better guidelines for making such decisions. This study examines the effects of two forms of customer involvement on new product outcomes: the traditional form of customer involvement as an information source (CIS) and the more active form of customer involvement as co‐developers (CIC). The authors offer a better understanding of whether customer involvement can lead to successful innovation by (1) identifying conditions that impact the effects of CIS and CIC on NPD outcomes, (2) contrasting the conditional effects of CIS and CIC to understand how they influence NPD outcomes differently, and (3) examining the potential substitutive relationship between CIS and CIC to understand their joint effects in improving innovation. They find that an experimental NPD approach that emphasizes trial and error learning moderates the relationship between customer involvement and new product outcomes. Specifically, the results reveal contrasting contingent effects of CIS and CIC: CIS is more beneficial for new product outcomes when firms take a more experimental NPD approach, whereas the effect of CIC is stronger when the NPD process is characterized with lower experimentation. CIS and CIC also substitute for each other in their contribution to new product outcomes. These findings suggest that each of the two forms of customer involvement has its unique advantages and is suitable for different conditions. When considering the adoption of CIC, firms should take into account their learning approaches as well as the effectiveness of CIS in the NPD process.