Abstract
As firms increasingly seek out consumers' ideas in various domains, they will encounter individuals with different levels of domain-specific knowledge. While both low- and high-knowledge consumers may be willing to share their ideas benevolently, the performance of the former is likely to be hindered by their lack of relevant knowledge in the problem domain. It is also well established that, despite their abundant knowledge, high-knowledge consumers may not perform in accordance with their full potential (due to factors such as shallow processing and inattention). We propose that firms can greatly enhance consumer performance in idea generation by customizing the task structure to overcome the distinctive challenges faced by these two types of consumers. In particular, we posit that the presence of stimulus ideas will greatly mitigate the lack-of-knowledge obstacle faced by low-knowledge consumers, and problem decomposition will enable high-knowledge consumers to perform more in accordance with their greater potential. We further examine the interplay between these two manipulations and outline a process for developing a customized system based on each consumer's knowledge level. Our hypotheses are supported by an extensive empirical investigation involving data from over 3,000 respondents across three studies. This research is among the first attempts to demonstrate that a customized idea generation system can be highly beneficial managerially.