Impacts of Accelerator Cohort Composition on Startup Performance

How does cohort composition affect the outcomes of accelerator participants? Literature on inter-firm and peer learning in entrepreneurship, provides mixed predictions: seeking diverse knowledge and being different vs. pursuing fast learning via similar peers. To reconcile the arguments, we investigate the key contingencies, types of knowledge transferred (market knowledge vs. technology) and the unique nature of competition in each knowledge type. Using data from accelerators and their participants during 2005-2018, we find an inverted-U-shaped relationship between market knowledge similarity in a cohort and post-accelerator performance and a positive relationship between technology similarity and performance. Our study reveals a peer learning mechanism in accelerators and provides a nuanced understanding of the effect of knowledge diversity on startup performance.