On Testing Hypotheses

What do venture investors, marketers, marketplace sellers and other professionals who work with probabilities all have in common?

They are masters at testing hypotheses — and at ruthlessly dropping the ones that don’t pan out. These people know how to part ways, at the right moment, with the startups, ideas or people that didn’t reach the goals they were set.

One of the most respected venture funds, Andreessen Horowitz, even published an article: “It’s Never Too Early to Fire… your CEO.” A stance like that clears away a lot of false expectations about how soft business is supposed to be.

Emotional involvement is the main enemy of rational decision-making. Founders try to pull investors into sympathy — for them it’s a fight for survival, after all. But the investor himself can’t afford the luxury of acting on emotion. Unless, of course, it’s a hobby, a way to have fun regardless of how the project ends. The temptation to believe you can have both — the result and the joy of the process — is strong, but almost always a delusion.

I’ve made the mistake myself more than once: I invested in a project on the strength of a business plan, and the real numbers didn’t match the forecast. Together with the other investors, we’d look for excuses, telling ourselves that one more round would fix everything. It’s hard to say no when people who poured their souls into a company come asking for help with tears in their eyes. But the truth is simple: the hypothesis didn’t work. Period. Better to recover what you can and write off the loss, so you don’t lose even more. Failure is a natural part of the road to success.

Trying to force a failed hypothesis to work is a futile attempt to prove your point — to the universe, or to yourself. It doesn’t work that way.

One of my more recent lessons was a food-production project that, in my view, still hasn’t been matched on taste — but the market didn’t take it. We kept hoping and dragged it out far too long. In the end we had to shut the production down and let the whole team go. The needless suffering and losses could have been avoided if we’d made the decision sooner.

On the other side: if a project is showing excellent results, it’s worth putting more into it — time, money, connections. A lot of people lock in profit and “exit” too quickly. But successful hypotheses are always far rarer than failed ones. From my experience, roughly one in ten really takes off. Exit too early, and that one success won’t cover the losses from all the rest. So it’s worth giving a project time to grow, helping it along, and not tormenting yourself with extra doubts.

The next levels of mastery include: — being able to test more hypotheses for less, lowering the potential losses; — clear criteria for judging what counts as success and what counts as failure; — and a plan of action for the case of failure, followed without flinching the moment the matching “trigger” fires.

Look at it philosophically, and our whole life is one continuous test of hypotheses: new opportunities, new acquaintances, relationships, careers, businesses, travels.

Knowing how to back the ideas that worked, and to part painlessly with the ones that turned out wrong, makes life more orderly, takes the edge off anxiety, and lets us walk our own path while keeping our own identity. What worked for others won’t necessarily fit us. And that’s absolutely normal!

Here’s to testing our hypotheses, growing the ones that work, and letting go lightly of the ones that don’t! 😎