June 27, 2012
I’m deep in the midst of a search for a partner to help me build the next chapter at High Peaks. It’s been a fun and fascinating process. Since starting the process with a very public casting call a couple months ago, I get asked almost daily by a friend or colleague “so how’s the search going?”
I generally reply, “I’ve met a bunch of people I really like and am having a lot of interesting conversations, but I have no idea how it’s going.”
“Yeah, I don’t really know. I’ve got a handful of prospects I’m very excited about, but it’s impossible to say how it’s going overall. After all, when you really think it through, at this point each one of them is more likely to not work out than get all the way to the altar.”
I’m actually really excited and optimistic about the outcome of my search. But I also know the importance of being realistic about any individual candidate.
The thing is, most people don’t really understand the way probabilities work. They have a couple of good conversations, think things feel good, and get overly excited. But outcomes are generally dependent on multiple variables, and multivariate probabilities are deceiving. They’re multiplicative, and that leads people to consistently, and often dangerously, overestimate the likelihood of a given outcome.
Let’s imagine I was pretty far along in discussions with a particular candidate and things were feeling good. Is that person likely to become my partner? The reality is that even if things are looking quite good, the answer is probably no.
Any outcome like candidate X becoming my partner is dependent on multiple variables, each of which has its own probability. Again, probabilities multiply. And when they do, things become unlikely to happen surprisingly fast.
Let’s consider a discussion with hypothetical candidate Sally. What has to happen for Sally to become my partner? There are at least three major elements, each with their own probabilities
- I have to decide I want Sally to be my partner
- Sally has to decide she wants to work with me
- Sally and I have to negotiate terms for her joining the firm that work for both of us
Pretty straightforward. But now let’s do what most people fail to do in assessing outcomes – think through the individual probability of these three independent variables:
- Based on our discussions thusfar, I think there’s a 2/3 chance that I’m going to decide Sally is the one.
- Sally has indicated that she’s quite interested in the position, but she would also be leaving a pretty good current gig, so I’d say she’s no more than 2/3 likely to say yes if I ask her to join me.
- If we both want to do it, I think we’re pretty reasonable people, so it’s highly likely that we’d be able to find terms that work. 90%.
So is Sally likely to become my partner? I doubt it.
It’s amazing how human psychology works – the vast majority of us are natural optimists, it seems. Most people hear 2/3, 2/3, and 9/10 and get excited, thinking it’s probably about 2/3 likely to happen.
But when you do the math and actually multiply out these probabilities – multiplying .67 x .67 x .9 – you end up at 40%. There is in fact a six out of ten chance that Sally will NOT become my partner.
It’s a surprising and important reality. In fact, in any three variable probability tree all you need is a 79% probability for each variable and you dip below 50% likely to achieve the final outcome. Sobering, isn’t it?
Of course this sort of math is unrealistically precise for the purpose of definitive, concrete analysis of any real world scenario. We never actually know the exact probability of any individual variable, after all. But the point is not in the precision, it’s in making sure that you understand how probabilities interact to effect outcomes in complex decision processes, even when you are evaluating things at a very rough level.
Entrepreneurs are by their nature optimists. It’s an absolutely essential component of their success. But I’ve seen that same optimism all but bring down a few companies in the past as entrepreneurs bet the ranch on outcomes they believe are going to emerge but which they can’t ultimately control. It happens with big partnerships, estimating product shipments, and assessing sales pipelines. Very painfully, it happens all the time around critical hires, as managers fall in love with one candidate, thinking that individual is going to join the company and solve all their problems. They fail to cover themselves by maintaining alternative candidates, and when the lead guy drops out of the running the company is, at least temporarily, up a creek.
So I simply ask you to remember a little bit about the laws of probability when making important planning decisions. Keep multiple candidates engaged in any hiring process. Don’t bet the farm on that big partnership with Apple/Google/Whomever. Remember that for it to come to be you’ve got to decide it makes sense for you, they have to decide they want to do it, and then you need to find mutually acceptable terms.
Even if those pencil out at 90%, 90%, and 90% you’re still <75% likely to get there. Is it probably going to happen? Sure. But is it so in the bag that you want to bet the company on it? I sure hope not.