Disclaimer: I work at Strive where we run management workshops, including ones around evidence-based interviewing. If you're interested in learning more, reach

Disclaimer: I work at Strive where we run management workshops, including ones around evidence-based interviewing. If you're interested in learning more, reach out to me at jesse@strive.co! An evidence-based approach (aka the scientific method) + a little knowledge about cognitive biases shows why questions like "Where Do You See Yourself In Five Years?" are not great. All the research into interviewing and cognitive biases says that asking questions for "color" reduces interviewer's ability to predict outcomes while simultaneously increasing their confidence that they've gleaned information from the interview. There are experiments where interviewers are given some diagnostic information (e.g., hard facts about background) and conduct closed-form interviews (yes/no, this/that questions) with the interviewees responding randomly. They'll still believe they've "learned more" about the candidate than they did from the candidates who they had biographical information alone. See the Dilution Effect (https://psychology.iresearchnet.com/social-psychology/social... https://psychology.iresearchnet.com/social-psychology/social...) and https://en.wikipedia.org/wiki/Sensemaking https://en.wikipedia.org/wiki/Sensemaking. Here's a fun paper: http://journal.sjdm.org/12/121130a/jdm121130a.html http://journal.sjdm.org/12/121130a/jdm121130a.html A good interview process reduces an interviewer's exposure to non-diagnostic information and directs their attention so they only evaluate diagnostic information (and does so in a consistent way). This is the basis for the "structured interview". "But Jesse," you say, "'Where will you be in 5 years?' seems diagnostic to me!" Questions aren't diagnostic; responses are! So you have to think about the nature of the responses and how those responses will be evaluated. Here are some alternate questions. Consider the responses you might get to these vs. the 5-year one. Are those responses more diagnostic? (Note: I just thought of these off the top of my head and haven't fully evaluated them). - At your last job, what career goals did you have when you joined? Did you achieve them? How? - Imagine you've been working here for 18 months. What would need to happen between now and then for you to not feel disappointed you joined? - Think back to a manager who helped you define and achieve a professional goal (big or small). What was that goal? How did they help you achieve it? Why did that work? As a first pass, these questions control for more and are likely to get higher-signal responses that are easier to evaluate in a consistent, interpretable way. How can you tell the original 5-year question isn't great, though? Think like a scientist. Why 5 years? Why not 1? Why not 10? 2? 4? How might those numbers change the answers you receive? Are you even testing to see how people's responses change as a function of that number? Are you even coding the responses in a way that would be able to detect how their responses change in response to that number? Is the variance in the responses small enough that you can even come up with a way to code them? For example, if the interviewee is startup savvy and you ask the 5-year question then they understand the vesting cliffs are at play. Suddenly you're taking into account how savvy the interviewee is about the way vesting works. Was that the intention? If not: more variance! Is it clear what the question is trying to get at? Do all candidates understand what it's trying to get at? Could candidate misinterpret the intent of the question given the fact that they're in an interview? Could different interviewers interpret the same response differently? It's a "bad" question because it controls for very little and therefore inherently produces non-diagnostic information (such is the nature of humans). The level of variance in all the components of this question is exactly the surface area on which cognitive biases become the determining factor. And the nature of cognitive biases is that you become more confident when they're operational, not less. The upshot for startups is that implementing a structured process is easier than doing so at a larger company. Half (or more) of the challenge of implementing a structured process is the package + selling + training + rolling out + reinforcement of the "new" approach. At a startup, you can get everyone on the same page much more quickly and bake it into the culture from the start (or close to it).