No, this blog-post will not be about Skynet becoming self-aware, and starting a war against the human race. It will be about using data in various HR processes.
Over the past six months I have “had the pleasure” of conducting a good deal of job interviews for an open position in our team of webinar producers. It's not that I don't like interviewing people, I simply can't convince myself that based on a 1-hour interview along with some additional practical tasks we will be able to determine if a person will be a good long-term fit for us. We found a couple of promising candidates, but it has not worked well so far with them; we are really struggling to fill the open position. The basic question which I started to think about recently is “what is wrong with traditional interviewing in general, and what is the solution?”
What is wrong?
- Interviews are incredibly subjective.
- Interviewers have the tendency to favor applicants who are similar to them (education, thinking, opinions and even appearance).
- Interviewees behave very differently at interviews than in normal situations.
- Not only do they behave differently, but they have also the tendency not to answer direct questions like “would you like to have this job” honestly.
- Interviewers have a set of skills in mind they are looking for; however, the required skills are usually based only on guessing and estimation. So they can't even be 100% sure that they are looking for the best possible skill set for the job.
- Possibly someone who is shy or nervous, who would otherwise be an excellent fit for the job, may deliver a poor performance at the interview due to her/his personal attributes.
Obviously the solution is eliminating the imperfect human element from the equation. I've read dozens of articles from or about celebrity CEOs and superstar managers describing the one time they hired a guy who was not a perfect fit for the job, but somehow he became a top performer – all of this based on an excellent hunch on the part of the interviewer. I am sure there are executives with vast experience who are able to assess a person based on a short conversation; however, I am also sure that in most cases they are just lucky to the above mentioned “not a perfect fit for the job” guy. There also must have been many situations when they hired someone based on a hunch, and he was a disaster.
Therefore the solution that looks viable to me is using data analysis in the assessment of job applicants; you might have heard about it as: Big Data. I know a bit about data analysis, and the best thing you can do with data is analyse the relationship between two or more variables. In the case of recruitment, the primary goal is to find out who the best employees are, what features they have in common. We can also use advanced statistical methods to determine the most important reason for high turnover, or which employees are satisfied and why.
I am convinced that HR right now has a good chance to lose the image of a department with no or very little hard business evidence. HR can use a variety of unused data to find patterns about their best employees – about engagement, satisfaction, turnover, etc.
If you think your organization is not ready, too small, or does not have the money to analyse data, you might need more information. You can check back in a couple of weeks at our website http://www.kakushinwebinars.com/ and see our upcoming webinar about predictive HR analytics.