Why co-founders fall out, & why success makes it worse.

As founders, we like to believe that if their company ever fails, it will be for an honourable reason. The market turned. A competitor got there first. The money ran out.

Noam Wasserman spent years at Harvard Business School studying what actually happens to young companies, tracking roughly 6,000 start-ups and 16,000 founders. His conclusion, set out in The Founder's Dilemmas, is that 65% of high-potential start-ups fail because of conflict inside the founding and leadership team. The largest single risk to the company sits in the room where the big decisions get made, in the working relationship between the people who started the thing.

The second finding is that this risk grows as the company succeeds. A start-up fighting to survive has no spare capacity for a power struggle, everyone is bailing water and fixing the printer. A scale-up has money, ambiguity, and several competing visions of what the company should become, which gives people a great deal more to disagree about. Wasserman's data shows that by the time ventures were three years old, half of the founders were no longer the chief executive, and when founder-CEOs did leave, 80% were pushed rather than choosing to go.

Twitter is the version most of us know. Four people had a hand in founding it, and over the next few years they took turns forcing each other out. Noah Glass, who came up with the name and built the early product, was the first to go. Jack Dorsey was the first chief executive, and in 2008 his co-founder Ev Williams removed him from the job and took it himself. Williams then lost the top job two years later, with Dorsey, by now the chairman, part of the manoeuvring that pushed him out. Dorsey eventually returned as chief executive in 2015. None of this was caused by a weak product. The company nearly destroyed itself over who got to run it

The easy explanation is personality: he is difficult, she is territorial, we stopped getting on. Personality is rarely the real fault line although on the surface it may appear to be. The fault line is structure: roles nobody defined, decisions nobody assigned, a division of the company that everyone assumed and no one ever agreed out loud. At three people the vagueness costs nothing. At thirty it is the thing splitting the leadership team in two.

You cannot reduce a risk you refuse to look at. Most companies pour their diligence into the market and leave the founding relationship completely unexamined, which happens to be the one variable the research says is most likely to kill them. The rest of this series is about looking at it.

Sources: Noam Wasserman, The Founder's Dilemmas (Princeton University Press, 2012); "The Founder's Dilemma," Harvard Business Review, Feb 2008. Twitter's founding history is documented in Nick Bilton, Hatching Twitter (2013).