Can Algorithms Replace Your Managers? Humans Need Not Apply...

Aug. 1, 2016

In the Headlines
man and woman shaking hands

A short, and so far undisputed, article made its way into the May 2014 issue of Harvard Business Review. In this piece, three well-respected psychology professors at the University of Minnesota argue that managers should leave recruiting to algorithms who outperform humans in the selection process.

Our analysis of 17 studies of applicant
evaluations shows that a simple equation
outperforms human decisions by at least 25%.

The analysis below taken from the article itself shows how algorithms are much better are picking the right candidates in a number of areas (supervisors' ratings of performance, advancement, and training performance).

Um... I can see many of you shaking your heads at the prospect that your next recruit AND your colleague for the foreseeable future can or should be selected by an algorithm. Simplistic and controversial all at once? I hear you.

Granted, the researchers caution against letting an algorithm make all the decisions and recommend using an algorithm to sieve through the initial candidates and then have the recruiting managers pick from the top three.

I read and re-read the article a bunch of times and I am still unresolved in my mind as to whether this is a good idea! I certainly have a bunch of questions. Here's some:

  • Can data really replace humans to such extent?
  • What if, as the hiring manager, I do not like any of the three candidates?
  • What if the chosen candidate turned out to be flawed and not right for the role?
  • To what extent would this process considered to be fair?
  • Would this approach be equally applicable to the internal promotion process?

My feeling is that both the idea proposed by the study and the questions I am raising will simply go away. I think both Big Data and Artificial Intelligence will make this an even bigger issue. All that I am left thinking is that as long as managers will drive an organization, their intuition will continue to trump algorithms. In the interim, what we need to teach them is to how to make better decisions in the interest of all those involved.

What do you think?

Image courtesy of Pexels