A predictably successful idea is composed of known unknowns that that a team can execute against.

A predictably successful company say two things: “if we can do this definite and precise thing, we will win” and “we are the team that can execute on that thing and here are the reasons why.” That thing can be extremely hard and risky, but it needs to be definite and precise.

A definite idea has a probabilistic roadmap composed of pieces that don’t have obvious unknown unknowns.1 “We will gain customers through our amazing UX” is not definite. “We will gain customers by building a product which is 5x more efficient than anything on the market” is definite.

A precise idea has a roadmap composed of pieces with clear scope and boundaries. “We will build an AI” is not precise. “We will build a deep-learning based system trained on user inputs and the outputs of mechanical turks we will hire” is precise.

SpaceX is an example of a predictably successful business - “if we can bring the price of a launch down to $75/kg we will win.” Instagram, while incredibly successful, is not a predictably successful business. Among all the photo sharing apps, there was not a definite and precise thing that they knew a priori would lead to completely dominating the market.

When dealing with a predictably successful idea, you should be able to say with confidence “if some controllable factors work out as we predict, then this team has a real shot at winning.”

Another way to think about predictably successful ideas are those that fall into the definite optimistic quadrant in the Peter Thielien sense.

The concept of predictably successful ideas was introduced to me by Matt Clifford in our conversation on the Idea Machines Podcast. While it’s meant to be a framework for thinking about startup companies, I think it can be extended to many other ideas as well.

1 Known unknowns are probabilistic variables with known distributions while unknown unknowns have unknown distributions.↩