Five Things That Make a Great Interview

(And Why AI Is Better At All of Them)

|

Adam Friedman

Founder & CEO

Blog Post: Five Things That Make a Great Interview (And Why AI Is Better At All of Them)

Ask a good interviewer what makes a great conversation, and they'll give you roughly the same answer: it's about making the other person comfortable enough to say something true.

That’s not as simple as it sounds. Most research interviews fail this test before the first question.

What a Great Interview Actually Requires

Journalists and researchers have studied this for decades, and the consensus is clear: a great interview requires active listening. It’s not just about waiting for the expert to stop talking so you can ask your next question. It’s about actually tracking what they say and following the thread when something interesting surfaces. It’s about being consistent, ensuring the energy and pace of a specific conversation doesn’t bias results. And it requires a certain psychological safety; the expert has to believe they can say something unpopular without it being held against them.

None of this is controversial. What is controversial, apparently, is admitting that the legacy expert network model is structurally bad at all three.

The Problem With Human Interviewers

The analyst on your typical expert call is good at their job. They prepared. They showed up on time. They even read the briefing document. But they're also taking notes (I used to transcribe calls live, so add hand pain to the cognitive load) or watching the clock, aware that this call is being billed by the minute. Active listening is a cognitive luxury researchers just don’t get.

Consistency is also. Change the interviewer and you change the data. What one person follows up on, another lets pass. The tone shifts, silences land differently, and you end up struggling to compare data cross interviews because the independent variable isn't just the expert, it's whoever happened to conduct the call that day.

And how psychologically safe can you really feel with a stranger? It's genuinely hard to say something critical about an industry when you're still wondering which side that stranger represents. This isn’t to say experts lie. But like all human interactions, these calls become a performance. There's a version of themselves they present to a professional context, and it's usually a little more careful, a little less specific, than what they'd say to a friend.

Five Things AI Does Better

This is where I might annoy some people (sorry, free rage-bait): AI moderators are structurally better at the core mechanics of a good interview than a human analyst. Not because they’re smarter. They’re not. But they’re consistent, and they don’t have bad days.

  1. AI is available when the expert is. Scheduling sounds like logistics, but it's actually a data quality issue. An expert who fit your interview between meetings, and who’s already running behind, gives you a different interview than one who sat down when they were ready. Async AI interviews happen on the expert's schedule, not yours.

  1. It removes social pressure. There's no human on the other end of an AI interview reading the expert's tone, reacting to their hesitations, or — consciously or not — signaling approval and disapproval through the conversation. Experts are more likely to say what they actually think when they're not managing someone else's reaction to it.

  1. It pays immediately. This sounds like an administrative detail. It isn't. An expert who isn't sure when — or whether — they'll get paid is an expert who's already a little resentful before the interview starts. Payment within 24 hours changes that relationship entirely.

  1. It's consistent. The same interview, conducted with the same energy, at two in the afternoon or eleven at night, for the third expert or the thirtieth. The data is comparable because the instrument doesn't vary. This matters more than most researchers admit.

  1. An AI moderator isn't splitting attention between note-taking, time management, and client expectations. It’s listening, processing answers and responding to what was said. When an expert mentions something unexpected, it follows up. Every time.

A great interview isn't magic. Get the conditions right and experts tell you what they actually know. That's the whole game.

If you'd like to see what this looks like in practice, reach out below.

Let's work together!

We're a team of people (and AI agents) committed to making high-quality primary research faster, more convenient, and, of course, more forward-looking.

Let's work together!

We're a team of people (and AI agents) committed to making high-quality primary research faster, more convenient, and, of course, more forward-looking.

Let's work together!

We're a team of people (and AI agents) committed to making high-quality primary research faster, more convenient, and, of course, more forward-looking.