Getting a job as a McKinsey consultant is hard by any measure. First you generally need to get into one of the best schools in the world. Then you need to ace a series of case interviews. These are not the brain-teasers you hear about in the popular press, like “How many golf balls would fit in this room?” Instead you will be asked a three-to-five part question along the lines of: This fictional airline is looking to expand into Moldova. How would you go about thinking of whether that is a good idea? OK. Here are a bunch of numbers, now what do you think? OK. Now that we’ve agreed it’s a good idea, here are the current schedules of all flights in and out of Moldova. Walk me through how you would create a hub-and-spoke schedule that meets the latent demand. OK. And so on. Oh. And the people interviewing have zero experience in schedule design, the airline industry or anything at all related to the case (The guy who knows airlines gets the case on setting up a churn reduction program for a Latin American telecom company).

Did I mention you need to ace the case? Did I mention you will go through at least 6 cases, and you have to ace all of them?

And what is usually overlooked is that about half the people that ace the cases still don’t get job offers because they fail to ace the second part of the interview that far more subtly probes them on how they handled difficult situations in their past. McKinsey has a rubric on how you should deal with a boss who is making poor decisions or how you get a diverse group to move in a direction you need them to or how you deal with a big mistake.

So, we can start with the premise that (1) It is hard to get a job at McKinsey, and (2) McKinsey, because it’s only real value beyond its brand is its people, spends a lot of time trying to make sure their interviews get them great employees.

Everyone who interviews at McKinsey ends up with a score. The highest scores get jobs, but there is still a significant range. Some people get jobs ‘on the margin’. There may even be disagreement in the decision room, but somehow they ‘sneak’ through. Others are so clearly McKinsey caliber that they walk through the decision process without a peep.

This scoring doesn’t end there. On every study, and more formally every six months, you get scored again. And again six months after that. You get a lot of numbers next to your name at the firm.

And this is where we run into the problem.

There is ZERO correlation between the scores people get on the interviews and the scores they get as employees. Zero. Nothing. Nadda.

How could that be?

McKinsey spends more time that just about anyone trying to hire the right people. They have a standardized process. The interviews are not “This candidate had a goodhandshake,” they are “This candidate was able to structure a problem and come up with a great answer in front of me with polish presentation skills. Just like they will need to do on the job.”

And yet…

It doesn’t seem to matter.

If someone kills it in the interviews or someone just sneaks in – they are both equally likely to make partner 6 years later.

Does that mean the interviews are a waste of time? Could McKinsey just hire people randomly and do just as well – and save all that effort.

I don’t think so. It doesn’t pass the sniff test.

Here is what I think is going on:

McKinsey recruiting does NOT find the best candidates. It just eliminates the worst candidates.

By only recruiting at top schools they have already had others do some screening for them. Then they make people do case interviews – which scares many others away from even applying. Of those left, the ones that don’t do well on the cases are eliminated. Many of those people would do very well if they had got jobs, but that’s not the idea. The idea is to eliminate the ones that will fail. Thy can live with false positives. So if you can’t do well on a case, you are gone. If you can’t sit in a room with a partner and come across professional you are gone. If you don’t have a track record of success behind you, you are gone. If how you handled things in the past doesn’t follow the ‘McKinsey way’, you are gone.

What you are left with is a group of people that are likely to be able to do OK or better at McKinsey. And that’s all you need.

What does this have to do with Marketing?

My theory is that we have the same problem in marketing. We have a bunch of ideas that we think might create value. We do them and some success and some fail. My theory is that we are no better at picking the most effective activities than McKinsey is at picking the most effective consultants. The best we can do is eliminate the candidates that will likely fail, and what we are left with is a bunch of candidates that are all about equally likely to succeed (or fail).

We are skilled enough to eliminate the bad, even if we aren’t good enough to pick the best.

Once you get this idea internalized you can see it everywhere.

Kickers in Football have zero correlation in relative ability from season to season. So if you have a choice between hiring the kicker who was #1 last season or the one who was the worst in the league, you might as well take the worst in the league. Obviously the worst NFL kicker is a lot better than a man off the street (say, me…), but it turns out he’s no better than the best in the NFL. The NFL does a great job of eliminating the people who can’t do it, but a terrible job of figuring out who is the best.

Investment Advisors also have zero correlation in returns from one year to the next. Well almost. If an investor is in the bottom 10% in a year he has a much higher than 10% chance of being in the bottom 10% next year. So we can avoid investing with him. But after that there is no difference in your expected return if you invest in a guy in the lower half of returns last year and the guy who was #1 in the country. We have the ability to eliminate the bad, but not the ability to pick the good.

Can you think of other examples?

How this affects your life

It should de-stress your life. Instead of spending calories trying to come up with the best idea or hire the best candidate, you can take solace in the concept that you just need to eliminate the bad ideas or bad candidates. As a manager, you should help your team eliminate things that are unlikely to work, but if seems ‘good enough’ to you, let them run with it. Their enthusiasm and engagement are likely to be far more important than what you think is ‘the best’ idea. Remember: You can’t predict the best. You are bad at it. You just didn’t know it.

It also means that the guy who created the amazing viral video is likely not a viral video expert. He’s likely ‘good enough’, but unlikely to be much better at creating the next viral video than someone else who is ‘good enough’. It’s a dangerous thought. Especially for those of us that are experts in our field with a track record of success.