Idea Implementation and Recommendation in Open Innovation Platforms

Abstract

Collaborative crowdsourcing communities help firms obtain ideas generated by the public sector at a lower cost compared to those generated in-house. However, the growth of these communities has led to a large influx of ideas. This has made it difficult for firm experts to identify and evaluate ideas and determine which ones to implement. We present a model to pre-evaluate ideas and to recommend those that are most likely to be implemented in the future. We use an elaboration likelihood model to integrate central and peripheral route processing by firm experts in idea implementation in relation to the content characteristics and popularity of the ideas. The empirical results based on 89,906 ideas submitted by 53,836 users to the new function discussion forum hosted by Xiaomi, one of the world’s largest electronics company, suggest that original, topic dispersion-focused, concrete, and negative ideas are more likely to be adopted and implemented. Ideas with lower number of views and ideas with higher number of comments or higher ratings are more likely to be implemented in the community. We implemented machine learning models for recommending ideas based on our empirical findings. Specifically, the results show that our proposed factors can provide an extra significant contribution in predicting and recommending idea implementation. We conclude by discussing the theoretical and practical implications.

Keywords: idea implementation likelihood, pre-evaluation model, content characteristics, idea popularity, recommendation