Study: TechCrunch Disrupt videos show VCs focus more on potential gains when interviewing male founders, more on losses in interviews with female entrepreneurs
Context & Ripple Effects
This study adds a mechanism to a funding-gap story that has been building for years: earlier data showed women make up just 7% of partners at the top 100 VC firms while all-female founding teams took roughly 2.2% of the $85B VCs invested in 2017. What was missing was evidence of how that gap gets produced inside the room.
By coding actual TechCrunch Disrupt interview videos, the researchers can point to questioning style itself — promotion-framed questions for men, prevention-framed for women — rather than inferring bias from outcomes alone. The finding also gives context to the parallel track of female-run venture funds positioning themselves as an alternative to the status quo.
First-order effects
- Female founders pitching VCs now have documented evidence that the framing of investor questions differs by gender, which changes how they can prepare for and evaluate pitch meetings.
- TechCrunch's Disrupt video archive becomes a research asset, tying the publication's own event footage to a debate about how its VC panelists conduct interviews.
Second-order effects
- If promotion-versus-prevention questioning translates into funding outcomes, it pressures firms to standardize diligence — and strengthens the case for female-run funds competing for deals the incumbent partnership structure underweights.
Third-order effects
- The pattern points toward accountability mechanisms for allocation: if later Crunchbase data shows female-only founding teams still capturing around 2.3% of VC funding years after these findings, question-framing studies become part of the evidence base LPs and regulators use to judge whether the gap is structural rather than meritocratic.
The trend: Venture capital is moving from anecdote toward measured evidence of gendered evaluation, as longitudinal funding-share data keeps the question of who gets funded open long after each individual study lands.