Beyond the Cohort: A Randomized Evaluation of Continuity Mentoring in Accelerators

Partner with multiple accelerators to run a multi-site randomized controlled trial (RCT). Randomize startups into: (a) standard program, (b) extended mentor engagement for 12–18 months, and (c) extended engagement plus algorithmic mentor matching by expertise/network fit. Track funding, revenue, hiring, and follow-on investment quality for 2–3 years. This trial directly tests duration and match quality as causal mechanisms behind mentorship effectiveness. If continuity and better matchmaking drive outcomes, accelerators can reallocate resources from dense cohorts to fewer, deeper mentor relationships. It also operationalizes “mediation” between startups and institutions (corporates, universities) as a testable lever. The impact is a clean, causal blueprint for redesigning mentorship architectures—potentially shrinking the performance gap with angels and improving capital efficiency of accelerator models.

References:

1. A meta-analysis towards the effectiveness of startup accelerators. Nikolaus Seitz, Martina Buratti, Erik E. Lehmann, Julie Kurrle (2025). Journal of Technology Transfer.
2. Learning from mentors and their influence for the growth of startups: accelerators vs angels. Sreerag Rajan, Snehal Shetty, R. Sundaram (2025). Development and Learning in Organizations: an international journal.