How I Build and Run Behavioral Interviews

Behavioral interviews have a bad reputation in engineering culture — they are seen as either vapid ('tell me about a time you showed leadership') or gameable by anyone who reads a prep guide. Kuhn's essay is interesting because it locates the problem not in the format but in the prep that interviewers don't do. The failure mode he names is precise: interviewers who haven't thought ahead about what 'a great answer' looks like end up grading every candidate as 'weak hire' or 'weak no-hire' by default, and the interview becomes a ritual rather than a data-collection exercise.nnThe first key argument is the temporal inversion. Most interviewers decide what a good answer looks like only after hearing the candidate's answer, at which point they are cognitively anchored to whatever they just heard. Kuhn argues for building the rubric and follow-up question tree before the candidate arrives — before you know what they will say. Once you know what you are looking for and how to probe for it, you start asking the right third and fourth follow-up questions instead of letting the candidate's framing determine what gets explored.nnThe second and most practically useful move is the set of general follow-up types he recommends. Asking for a timeline consistently reveals things candidates have glossed over — weeks of inaction between noticing a problem and addressing it, for instance. Asking 'what was the outcome?' sometimes elicits admissions ('I eventually had to fire the person') that the candidate would not have volunteered. Asking 'how much did it improve and how do you know?' surfaces both the numeracy and the honesty gap — a head of technical recruiting who says 'our outbound response rate went from 11% to 15%' on a small sample is telling you something important about their relationship with data.nnWe picked this because clear thinking about people-evaluation is rare, and Kuhn writes unusually well about it from actual experience rather than theory — useful for anyone who hires people, is hiring their first people, or is preparing to be on the receiving end of these questions. The sobering implication is that if no one in your hiring process has spent two hours preparing the interview, the interview is probably not telling you what you think it is.