July 14, 2017

Algorithmic Hiring: What You Need to Know

The consumer goods company Unilever was recently in the news for their new hiring process based almost entirely on algorithms. Candidates are discovered through targeted ads on Facebook and career advice sites, and they only meet face-to-face with an interviewer after surviving an algorithm, completing a series of online games, and responding to a few questions on video.

Similar models have become all the rage within industrial / organizational psychology. Proponents of these systems say computers have found a way to understand human behavior through game theory and social media data that will revolutionize modern hiring practices.

There are plenty of potential benefits of a computer-based selection process, including greater speed, lower costs, and a reduction of the biases and bad interview practices research has shown humans engage in all the time. But can machines predict performance as good as or better than humans who use a traditional selection process? And is it worth the sacrifice of a personal approach?

As is the case with just about anything, the devil is in the details. Your success with algorithm-based systems will depend on:

  • The role itself
  • How much data you have
  • How consistent the predictor and outcome is over time

Yes, job-related games or social media data can predict job performance in some positions. But to implement a system like Unilever’s, an intensive amount of validation needs to be done to ensure the data applies to the position within YOUR organization. This could be a beneficial approach if you’re handling a large number of roles and candidates within a very consistent environment. If that’s not the case for your company, the expense and time to collect enough data to satisfy your lawyers won’t be worth it.

Luckily, there are other methods validated by industrial / organizational psychology that accurately predict the performance of your candidates without making them feel like part of a cattle call.

At TSP, we believe that textbook solutions only work within textbook organizations. In our executive assessment, we employ a personality assessment, behavioral interview, and role-play. That’s three different opportunities for a candidate to display their abilities within each competency. Most importantly, the personalization of the process allows for full agility of the criteria we are measuring.

In a previous blog post, I discussed tspFOCUS, which we use to assess candidates and create a customized interview guide. Understanding the approach behind why a candidate makes a decision during an assessment is almost as important as the decision itself. When the sample size of the employee group is too small, like the executives we put through our assessment process, the personalized approach is the only way to go. 

On paper, algorithmic hiring may look like a better mousetrap. But there are ways to put individuals through an advanced selection process that won’t sacrifice personal experience. The best hiring practices are as unique as the companies using them, and with a emphasis on the #powerofpersonal, TSP gives you the best shot at building your best team.

Tags: consulting & assessment, I/O psych

Marc Prine PhD

Marc is a Director in the Consulting & Assessment practice at Taylor Strategy Partners where he works with clients on improving their performance by using data to better select and develop their people. Dr. Prine earned his Ph.D. in Business Psychology from The Chicago School of Professional Psychology, M.A. from West Chester University of Pennsylvania, and undergraduate degree from Temple University. He is an adjunct professor in statistics and his work has been published in Forbes and FastCompany. In his spare time, Marc can be found looking for golf balls in tall grass, yelling as if the players on his fantasy team can hear him through the television or doing his best Andrew Zimmer impression and trying the most interesting thing on a menu.



More from this author