How I Failed Miserably as an Interviewer (and How You Won't)

At the beginning of the millennium, the startup I worked at was recruiting heavily across all the roles a tech company typically requires, and as part of my job, I was involved directly in the hiring process.

Recruiting Dilemma (Chat GPT)

The relevant part of the process usually went like this:

• The candidate waits in the reception area.

• I walk over, introduce myself, and chat briefly about trivial things (parking, coffee, restroom breaks) during the 15-45 second walk to the interview room.

• We start the formal interview.

• After the interview, I submit my feedback about the candidate.

• Repeat with the next candidate.

After several iterations, I noticed something troubling:

By the time we began the formal interview, I had already formed a strong initial opinion – positive or negative – about the candidate. The formal interview and subsequent assessments had to be exceptionally good or bad to alter this early impression.

Let me emphasize that:

Within just 15 to 45 seconds of meeting a candidate, I had already formed a lasting opinion. Even after an hour or two of deep-dive interviewing, my initial impression rarely changed.

This concerned me deeply because I couldn’t identify the bias clearly, and I worried about how it might affect the quality of our hires.

So, I began researching the topic with three critical questions in mind:

1. Was this behavior unique to me, or common among interviewers? (Suspicion: common)

2. Could this bias negatively impact our hiring quality? (Suspicion: definitely)

3. Assuming the answer to the second question is "Yes", how can I mitigate this risk?

I'll start with answering the third and most important question:

Always test your candidates (focusing on testing the skill and performance for the role they are applying to) —and make the test a major factor of the decision process.

Always find objective ways to measure candidate quality. Resumes, interviews, and past experiences are indicators – but they alone cannot reliably predict future performance. Tests and structured assessments are critical tools to complement your hiring process.

Here’s what my research uncovered:

• This behavior is not unique to me. Humans form snap judgments about others extremely quickly. A classic example is the influential 1993 paper by Ambady & Rosenthal, "Half a Minute: Predicting Teacher Evaluations From Thin Slices of Nonverbal Behavior and Physical Attractiveness."

  They found that people viewing video clips of teachers lasting just 2 to 10 seconds could accurately predict student evaluations given at the end of a semester. Participants' quick judgments based on nonverbal cues strongly aligned with opinions from students who had interacted with the teachers over an entire semester.
  

• Traditional interview and selection processes aren't highly effective at predicting actual job performance.

  A compelling example comes from a 1987 study published in the Journal of the American Medical Association. The study compared outcomes of two student groups at the University of Texas Medical School at Houston after a legislative mandate increased the class size from 150 to 200 students. The additional 50 students were initially lower-ranked and mostly rejected by other schools. Yet, surprisingly, both these students and the initially higher-ranked students showed similar graduation rates, academic performance, and honors, and eventual professional success.

In short, your initial impressions are probably – like mine – quick, powerful, and potentially misleading. To ensure you hire well, always rely on structured, objective tests alongside any other process you put in place for the hiring selection process.