Interview Scoring

Several years ago, I was involved in a lot of interviewing for several technical positions at the company where I then worked.  Generally, a candidate would come in, and multiple engineers would interview them individually for 30 to 60 minutes each over the course of the day, and then we'd get together in a conference room to discuss their merits.  I realized that it was helpful to have a quick, common starting point for the discussion.  If everyone agreed that someone was a dud, there was no point in continuing to talk about them.  Likewise, if everyone thought they were amazing, we would be motivated to get an offer to them quickly and try to lock them in before they got another offer elsewhere.  To facilitate this discussion, I developed a simple 1 to 10 point scale for broadly rating the viability of a candidate:

1. I will immediately tender my resignation if you offer this person a job.
2. Extremely weak candidate.  This person is likely to be a long-term drain on resources, and may well never make the transition to being a net positive contributor.
3. Weak candidate.  Initially, this candidate will be a net loss, requiring training, mentoring, constant guidance and close supervision.  They may grow in time, and become a net contributor, but it is not clear when (or if) this will occur.
4. Mediocre candidate: This person demonstrated some obvious deficiencies in the course of the interview, although they were not a complete train wreck.
5. Fair candidate:  This person showed no particular brilliance, but no glaring deficiencies either.  They will require support and assistance, but it is reasonable to expect that they will become a useful if uninspired contributor in time.
6. Reasonable candidate: This person seems like they may become a valuable team member, after some start-up delay.
7. Strong candidate:  This person is likely to become a valuable member of the team fairly quickly.
8. This is an extremely strong candidate.  While not a rock star, they have an impressive breadth and depth of knowledge, and will come up to speed very quickly.  Unless there is a rock star candidate in the line-up, we should very strongly consider making this person an offer.
9. This person is a rock star.  They will make an immediate and valuable contribution to the company, and we are unlikely to find a better fit for this job than this candidate.
10. I will immediately tender my resignation if you do not offer this person a job.

At the start of the meeting, everyone would, without discussion, secretly write their score for the candidate on a piece of paper.  Then we would reveal the scores all at the same time.  The results were usually quite interesting.

Often, we would find that one person would rate someone wildly differently, higher or lower, than everyone else did. If the candidate was in the acceptable band to most of the interviewers, then the discussion became an exploration of the cause of the disconnect in scoring.

Engineers being engineers, they would often add precision to the scale.  Scores of 7.5 or 6.5 were common.  And once the discussion was underway, people were free to re-evaluate their initial score in light of the persuasive arguments of the other interviewers.

Depending upon the position, the rate at which a company is growing, and the level of candidate sought, the threshold for further consideration may well slide up and down.  Some people will never consider anyone below an 8.  Some jobs may be perfectly well filled with a 6, but you want to at least know that you're getting a 6 and not an 8.  I would generally dismiss out of hand anyone below a 5;  the dot com days of filling chairs with any warm bodies to boost acquisition valuations are, I hope, a thing of the distant past.

Finally, the score provided a short-hand for the relative comparison of multiple candidates, often across multiple weeks of interviewing.

This is obviously just a first pass go/no-go filter and a starting point for discussion.  In some cases though, it helped keep the wrap-up meetings VERY brief!