What type of talent begets success?

It's not the identification of talent that matters

Malcolm Gladwell is at it again.  I can think of no one who has done more of a yeoman's job injecting ideas into the popular discussion.

Gladwell's latest offering asks us to consider the problem of talent selection.  

After an anecdote on the difficulty of picking NFL quarterbacks based on college performance, we get to the heart of the question: "There are certain jobs where almost nothing you can learn about candidates before they start predicts how they’ll do once they’re hired. So how do we know whom to choose in cases like that?"

But the NFL quarterback problem functions as a metaphor for the worldly problem Gladwell wants to tackle, which is the relationship between school teachers and the performance of students. 

“Eric Hanushek, an economist at Stanford, estimates that the students of a very bad teacher will learn, on average, half a year’s worth of material in one school year. The students in the class of a very good teacher will learn a year and a half’s worth of material. That difference amounts to a year’s worth of learning in a single year.”  Those are pretty compelling numbers.

There is also a brief treatment of the performance of financial advisors. But Gladwell wants to focus on teaching. 

How do we identify teaching talent?  How important is it?  How do we improve the quality of teaching for our children? This is a non-trivial problem, but the larger question of domain-specific talent looms in the margins. How do we identify domain specific talent?  Perhaps more importantly, how do we develop it?

A case that comes to mind is the question of talent in the software business. 

I have experienced enormous performance gaps between software engineers. You might have several engineers in a room who have equal education, equal experience, equal pay, and yet the performance gap between them can be astounding.  The gap can be as high as 10x.  Yes, that’s right.  You can have an order of magnitude difference in performance.  But that is not the hypothesis I want to define.

Here is the hypothesis: performance variation becomes relevant through the choices of the judge who determines what performance looks like.  In other words, it’s not the identification of talent that matters as much as the definition of criteria that determine talent measurement in the first place.

For better or for worse, the judge translates to the rest of us what performance looks like.  What are the measures?  What are the variables?  What are the indicators?  What is the shape of the target?  How we define talent up front is, I believe, the most important question we can ask ourselves here.  Who has the talent to define the success criteria for talent?  Perhaps more importantly, who has the power to define the nature of talent?

Some other examples might help.  1. Who determines the performance of a stem cell researcher? 2. Who determines the performance of a social worker?  3. Who determines the performance of a psychologist?

Let’s go back to software engineering for just a moment.  One might use productivity metrics to determine performance. 

Typical output measures might look like this: lines of code per day, number of defects per build, consistency of documentation, etc.  But these can become grossly misleading metrics.  They presume that more code and more documentation are inherently what you want.  But bigger software is not always better. In fact, bigger is often much, much worse. Thus the challenge of defining talent in software engineering remains with us still.

Then, somewhere in the middle of this piece, I had a flashback.  I was standing in the parking lot of an office building basement, a building I haunted during the darkest days and nights of 2002 and 2003.  The tech economy had not yet achieved rigor mortis -- the body on the table was still twitching with vestigial echoes of life.  A gruesome image for a gruesome time.

I was talking with one of my compatriots while waiting for the elevator.

“You see, in my opinion, we are not taught the one single skill we need in order to lead in the working world.  We are taught to succeed on our own, in a book, for an essay, for a test, a paper test, and working with others to get to an answer is considered cheating.  We are taught to compete with each other.  Grades are curved based on the standard deviation to separate mathematically the talented individuals from the other two groups: the mere mortals and the lunkheads.”

I was leading up to my grand point. "There’s really only one skill we need to succeed in an organization, one very basic skill. We need to be able to get a group of people to work together to agree on the problems we are trying to solve and then to set aside our egos to work together on the solution. 

That’s it.  That is all you need to know. If you can do this, you have already succeeded."

(Image source: Newlearningplaybook)