Are we doing the wrong thing by interviewing candidates?

I recently came across articles* suggesting that a job interview is too subjective to predict the performance of a future employee. Our unconscious biases would lead us to hire people who are like us, but not necessarily the most capable Professionals. As I did not love the idea of having spent the 10 last years of my life doing random interviews, we decided to have a closer look.

 Was there “really” so little correlation between our impression on a candidate during the job interview and their future performance as an employee?

We asked our clients two questions (**):

“Think of your 5 last hires:

-       On a scale from 0 to 10, how did you rate the candidate after the interview?

-       On a scale from 0 to 10, how would you assess their performance some months after the hire?”

•  How those who precisely predicted the future performance of the candidate achieved that?  

o  They conducted “structured interviews” (in a nutshell, this means they asked the same questions to all candidates. They asked questions that could be quantitatively analyzed and compared).

o  HR attended.

o  They used other HR tools (assessments, business cases, etc)

• Who were the candidates who over-performed?

o  It seems they were an introverted type of persons. Let us give them a slight premium when interviewing them!

• What happened to those who failed in assessing the candidate?

o  The hiring manager came in unprepared: he/she was overloaded, read the CV  ten minutes before, did not ask standard questions

o  The hiring manager thought he/she could rely on their gut feeling.

o  The definition of the role was not so precise, resulting in different interpretations and thus, different expectations. For example, the famous “… and Ad-Hoc projects”

o  The company was in a rush to hire. (“It is the 29th of the month, we must make an offer now”).

o  No onboarding. Typically, the manager is on parental leave, or constantly in meetings when the employee starts.

o  Something has changed in the life of the candidate: a divorce, burnout, etc.

 By the way, how would you assess the performance of your spouse in comparison to your first date? ?? 

Additional analysis for those who want to keep reading:

(*) When one starts searching on that topic, it appears a lot of studies have already been published.  I am sharing below two links which I found relevant, but they are plenty of others- way more complex and detailed:

The google research (2014)

Interview of Olivier Sibony, consultant at McKinsey (2021 - sorry, in French)

(**) Sample of 87 interviews, based on the declaration of the hiring managers. So, we get some first indication, but no exact statistics yet.

Qualitative interviews conducted to understand the biggest variations.

? About 3 recruitments out of 4 are a “success”: the interview gave a good prediction of the future performance of the candidate.

? That ratio seems to be higher in Switzerland than in France (my home country) or in the US (where most articles I found were published). My guess is:

-       In Switzerland, we have the work certificates system (Arbeitszeugnisse/ certificats de travail). Those certificates prevent hiring managers from misunderstanding the CVs of the candidates.

-       Switzerland is a relatively small country: we all know each other, right? ;-) Reputation is important here. Since we also have a low unemployment rate, I believe candidates tend to be themselves during interviews, and do not over push.

? We did not measure the performance of the candidates who did not get hired, of course. We all probably lost talents without realizing it!

? The measure of the performance of an employee is also a very subjective exercise, the same way we had biases when conducting an interview. It’s important to realize the performance of an employee depends also on the performance of the management.

? It seems no tool is perfect (assessment, business case, try and hire day, references), so I believe, we can keep our good old interviews, they are as valid as others. But let’s prepare them carefully!