In response to last week’s post showing raw data about the propensity of “idea people” to take the CEO position within their founding teams, Tadej commented:
I don’t think having the idea should generally affect your position in the company (that should be determined by your abilities and motivations) – some people say that having an idea is worth 20 bucks, the rest is implementation.
Tadej makes an excellent point about what teams “should” do. In practice, what do teams actually do?
To give us some deeper data regarding this, I ran a quick regression analysis to assess the strongest factors affecting which founder takes the CEO position. The factors I included in the regression were as follows:
- Serial entrepreneur — Has this founder already founded a prior company?
- Industry experience — Has this founder worked in the same industry before?
- Years of experience — This founder’s # of years of prior work experience at time of founding
- Was idea person — Did this founder come up with the idea on which the venture was based?
The first 3 are meant to capture the “abilities” factor mentioned by Tadej, with the fourth being the variable I described in the “Idea People” post. The dependent variable in the regression was whether the founder was the founder-CEO or not.
Quick technical notes: The pseudo-R-squared of this basic logit model was .07, the Prob>chi2 was .0000, and I clustered the standard errors by venture. Regarding the 3 abilities metrics, 26% of the founders had prior founding experience, 31% had worked in the same industry before, and the average prior work experience was 15.9 years.
Side note: Interestingly, the Idea Person variable was not significantly correlated with Serial Entrepreneur and with Years of Experience, but was significantly negatively correlated (r=-.22, p<.01) with Industry Experience.
The results were as follows:
- Serial entrepreneur: Strongly significant positive effect (p<.001) on becoming founder-CEO.
- Industry experience: No significant effect.
- Years of experience: No significant effect. (Also not significant when curvilinear relationship was tested.)
- Was idea person: Strongly significant positive effect (p<.001) on becoming founder-CEO.
So although role assignments “should” be done without regard to who was the idea person, in practice, who had the idea seems to have a strong effect on who becomes the founder-CEO. Why do you think we see this divergence between “theory” and practice?