Overview
This series asks whether startup accelerators create value or mainly select and signal promising ventures. It separates admission, program mechanisms, founder absorption, design, ecosystem alternatives, and measured outcomes, then tests them against a meta-analysis and applicant-based evidence. The overall finding is conditional: participation is modestly positively associated with performance, but effects vary widely, financial outcomes dominate, and causal shares remain unresolved. An accelerator’s practical value depends on the capability or relationship it adds relative to the venture’s best realistic alternative.
Learning outcomes
- Accelerators can create value through funding, learning, networks, and legitimacy while also selecting ventures with stronger starting prospects.
- Across 68 effect sizes, participation has a modest positive association with venture performance (r = 0.102), not a guaranteed causal return.
- Very high heterogeneity and publication-bias signals mean the average cannot stand in for every program, venture, or setting.
- The evidence is clearest for financial outcomes; durable strategic value remains less clearly established.
- Judge marginal value by the missing capability, realistic alternative, ownership and attention costs, outcome, and time horizon.
Keywords
Preview every chapter · Unlock full text, audio, and video from chapter 4
PlansFAQ
Do startup accelerators create value or mainly select strong founders?
They can do both: programs may create capabilities through resources, learning, networks, and legitimacy while selective admission favors ventures with stronger prospects.
How strong is the evidence that startup accelerators work?
The available meta-analysis finds a modest positive association with venture performance (r = 0.102 across 68 effect sizes), but high heterogeneity and nonrandom participation prevent treating it as a universal causal return.
How can a founder decide whether an accelerator is worth the cost?
A founder should compare the program’s specific offer with the venture’s best realistic alternative for its current bottleneck. Count equity, time, attention, and delay, and define the outcome and horizon in advance.
Why do startup accelerator results differ across programs and ventures?
Results differ because starting capabilities, founder and team fit, program design, ecosystem alternatives, selection, signaling, and outcome measures differ. A strong average therefore cannot predict the value of one program for one venture.