The best interview I ever had didn't feel like an interview. I sat down across from the hiring manager, she asked me what I was working on lately that I actually enjoyed, and we spent the next forty minutes in a conversation that drifted from a side project I'd been building into questions about how I'd approach a problem her team was actively stuck on.
What made it good was that she was clearly listening for information rather than running a checklist. She asked follow-up questions because she was curious, not because the script demanded them. When I said something she disagreed with, she said so directly, and we had a small, productive argument about it. By the end I had no idea whether I had "done well" — but I knew exactly what kind of work we'd be doing if I joined, and exactly what kind of person I'd be working for.
That's the test, I think. A good interview is one in which both sides leave with more accurate information about each other than they came in with. A bad interview is one in which both sides leave with a strong sense that a performance has been completed.
I got the job. I took the job. The work itself turned out to be exactly what the interview had implied. That's the rarest thing of all.
The Structured Interview Research
For decades, behavioral psychology has had a fairly consistent answer about what makes hiring decisions better: structured interviews, where every candidate is asked the same questions in the same order and rated against the same rubric, outperform unstructured interviews by a wide margin. The data is robust enough that the major consulting firms, the most selective universities, and most of the federal government have moved to structured formats.
What's interesting is that almost no candidates prefer structured interviews. They feel impersonal. They reward the type of person who can produce on-demand examples of "a time you led a team" without sounding rehearsed, which advantages a specific subset of candidates. The structured format produces better hiring outcomes on average, but it produces a worse candidate experience on average, and the tension between those two outcomes is a real and ongoing debate inside HR research.
What the Best Interview Actually Tests
The best interview I described in the original post wasn't structured in the technical sense, but it wasn't unstructured either. It was what researchers call a "work sample" interview: the interviewer brought a real problem the team was actively wrestling with, asked the candidate's approach, and engaged in a genuine discussion of the candidate's reasoning. Work-sample interviews are arguably the highest-validity interview format in the literature, and they're also the closest to what the actual job feels like.
The reason work samples outperform other formats is mechanical. You're observing the candidate doing some approximation of the job, rather than describing past instances where they did some related job. Description is heavily filtered through self-presentation; doing is much less so. Most candidates who can talk about their past work persuasively cannot actually execute the work in front of you with the same fluency, and watching them try is much more informative than watching them describe.
The Failure Modes
The most common failure mode in interviews is what I'd call "the impressed interviewer" trap. The candidate is articulate, has a strong narrative, looks the part, and the interviewer comes out of the conversation feeling positively without being able to articulate specifically why. This is, statistically, one of the biggest sources of bias in hiring. The decision feels like an assessment but is actually a vibes check, and vibes checks correlate poorly with actual job performance.
The corrective is to write down specifics during and immediately after the interview. Specific things the candidate said that you found insightful. Specific moments where they seemed to misunderstand or sidestep. Specific evidence of the skill the job requires. If the post-interview notes are mostly adjectives ("smart, articulate, energetic"), the interviewer was running on vibes. If the notes are mostly examples ("when I asked about scaling the team, they immediately distinguished onboarding velocity from contributor velocity, which I hadn't expected"), the interviewer was running on data.
The Signals Candidates Should Read
The other half of this is what candidates should look for in the interviewer. Good interviewers ask follow-up questions because they're genuinely curious, not because they're working through a checklist. They acknowledge interesting answers rather than just nodding. They volunteer real information about the role, including the parts that aren't great. They run slightly over time, in a good way, because the conversation got interesting.
Bad interviewers, almost universally, share a single trait: they don't seem to be listening to your actual answers. They're either running their next question silently in their head while you finish, or they're matching your responses against a template and looking for keywords. If a candidate gets a strong sense that the interviewer is bored, they should take that signal seriously. The interview is a two-way assessment. A bored interviewer is usually a bored colleague, and the role itself is unlikely to be more compelling than the conversation describing it.
What I Keep Coming Back To
The interview that got me to think most clearly about my own career was the one where the hiring manager spent fifteen minutes describing the worst part of the role before describing any of the good parts. It was disarming. It made every subsequent claim about the role's upside more credible. By the end of the conversation I understood the job in three dimensions instead of two, and I made the decision to take it on better information than I usually have.
I've tried to import that into my own interviews when I'm on the hiring side. It's harder than it sounds. There's institutional pressure to sell the role, and acknowledging the rough edges feels, in the moment, like you're undermining your own pitch. But every senior hire I've made who stayed and thrived was the result of an interview where I led with the hard parts. The candidates who can absorb the hard parts and still want the role are the ones worth hiring.