What's your Interview Score?

Let's play a game. You're a Hiring Manager (HM) and you're trying to hire for a Head of Marketing. You and 3 of your teammates will conduct 4 interviews, each specializing in a different theme. You'll then bring your notes together and make a decision. 

Notice first that 75% of the data and "reads" about this candidate will come from your teammates, not you.

If you are a control freak, or if you care about making great hires, that should concern you. How do you know if you can really trust that information?

The easy answer is that your company only hires great people, you're all smart, everyone has been trained in how to interview well, etc. There are a few problems here:

1. Most people aren't well trained at interviewing 
2. Many smart people are still bad interviewers
3. It's still very subjective so it's easy to fool yourself 
4. Feedback cycles (from bad hires) are long, with messy causation links. 

The result of this is it's very easy to convince yourself that you and your teammates have a better interviewing process than you do.

I've been playing around with the concept of an interview "score" that can be assigned to any interview transcript (the set of questions and answers). This is not the subjective judgment or analysis rendered by the interviewer—we're literally just talking about what was said during the interview.

How would we go about "scoring" an interview? Remember, an interview serves two primary functions:

1. Gather information that predicts whether or not the candidate will be successful in the given role
2.  Create a great candidate experience to attract/sell to the top-performing candidates that you want to hire.

Let's assume you have great rapport dialed in and ignore Category 2 for now. How would we judge whether or not you elicited predictive information? 

A great start would simply be counting the number of times the transcript shows the following:

1. A story from the candidate's past
2. With specific (ideally, verifiable) information and facts
3. About a relevant outcome or result
4. That the candidate was responsible for (causation)
5. Calibrated by a credible third party or benchmark (magnitude)
6. [Ideally] With an understanding of the root cause (deep motivation).

To be clear, this is asking a lot. Not every story will have this. Not every story from the candidates past will be important enough to go into quite this level of depth. But it's a good start.

You can think about going through a transcript and giving a point (or perhaps more) for every time the interviewer elicits a story that hits these key elements. How many quality, predictive stories did they get?

 And of course time is the limited resource in an interview, so a highly skilled interviewer knows when it's time to move on. You don't spend 15 minutes on one story getting every final detail. You also don't want to get a surface-level story that has no predictive value. 

This is the foundational layer of good interviewing. The predictive data. Without it, your discussions of candidates will be full of unsubstantiated speculation. Sure, people will have strong opinions but they will—mostly—not be justified (the data on hiring success rates proves this).

With a high amount of quality data, however, everything changes. Now we can start moving to the next layer of great hiring decisions: data interpretation. (For another article.)

Happy hiring!