Talk notes
The bus factor is an old concept in open-source software. How many contributors can be hit by a bus before the project becomes unmaintainable? In closed-source venture-backed startups, the question is the same with different stakes. How many engineers can leave before the company's engineering velocity collapses? The answer correlates surprisingly well with subsequent fundraise difficulty.
We define the bus factor formally as the Gini coefficient of commits across the top-10 contributors over the trailing 180 days. A Gini of 0 means perfectly equal distribution, every contributor commits the same volume. A Gini of 1 means one contributor commits everything. Empirically, organizations that sustain a Gini above 0.7 over a 180-day period have a 3.4× higher rate of subsequent fundraise difficulty (defined as a downround, an extension round, or a recap).
The signal is leading because contributor concentration is structural, it doesn't change quickly. By the time a fund's diligence team is reviewing the deck, the bus-factor risk has been in the data for 6-9 months. Most VCs are surprised when a company's lead engineer leaves and the engineering velocity falls off a cliff. The bus factor would have flagged the structural risk well before the resignation.
Sizing the signal across different total contributor counts is the practical challenge. A 5-person team and a 50-person team can both have a Gini of 0.7, but the operational implication is different. We control for this by reporting the Gini alongside the total contributor count, and by raising the false-positive flag for any org with fewer than 8 commit-authors over the trailing 180-day window.