Performance Evaluation of US Startup Accelerators Using Data Envelopment Analysis combined with Goal Programming Approach

Startup accelerator program is a new incubation model to accelerate venture creation providing specialized mentorship, networking opportunities and financing growth driven businesses. Since emergence of the first accelerator in US in 2005, the number of seed accelerating programs has quickly grown worldwide. Performance evaluation of these emerging programs not only could clarify relative success of this new incubation approach but also help out Iranian emerging accelerators to benchmark from pioneers of the industry. The evaluation could also assist startups and entrepreneurs to find out which of the existing accelerators are well-performed and thus ease decision making to select them as their incubator. So in this paper we aim to measure the performance of US top accelerators using a DEA method combined with goal programing approach. The inputs of the evaluation consist of numbers of mentors and startups, investment on each startup in average and in general, and the share of accelerator in startup company. The outputs in turn are different and include number and value of exits, and fund raising. Comparison of the final result with Seed Accelerator Ranking Project (SARP), shows different sorting and it is due the distinct evaluation approach. SARP considers only outputs of accelerators but we take in to account both outputs and inputs in the evaluation which could better illustrate the efficiency of DMUs.