Podcast interview questions template: 21 questions, three sections, and I never finish them

Sunny Ray · Mar 28, 2026 · 5 min read

My podcast interview template is 21 questions in three sections of seven: the guest's backstory, their aha moment, and their company and where it's going in five years. AI drafts the questions from research on the guest, the guest gets them in a shared doc before the interview and can edit them, and I almost never get through all 21. That's by design. The questions are artillery, not a script.

I walked a founder through this on March 27. He was building his own show end to end, campaign, booking, pre-qualifying call, taping, and asked where to focus. The interview structure was the part he hadn't thought about, and it's the part guests remember.

Three sections, seven questions each

Here's how I described it to him:

"For me, I do 21 questions, and each section has seven questions."

The number isn't sacred. You could do ten or a hundred. Twenty-one works for me because three sections of seven gives a clear shape to the conversation and more than enough material for the time.

The three sections run in order:

• Backstory. Where they grew up, what they did before, the moments that shaped them.
• The aha moment. The point where they saw the problem, or the opportunity, that became their work.
• The company and the future. What they're building now, the mission, and where they want it in five years.

The order matters. People open up when they talk about where they came from. By the time you get to the company, they're relaxed and speaking like themselves instead of reciting a pitch.

What the questions look like

A few examples of the kind of question in each section, before they're tailored to the guest:

• Backstory: where did you grow up, and what did you think you'd be doing at this age? What was the first job that taught you something you still use?
• Aha moment: when did you first see this problem up close? What made you think you were the person to fix it?
• The company: what does the business look like in five years if it goes right? What's the bet most people think is wrong?

The research is what turns those into questions only this guest could answer. "Where did you grow up" becomes a question about the specific city, the specific first job, the specific pivot. That's what makes a guest lean in.

Person over company

This is the part most business podcasts get wrong. They start with the company. "Tell us about what you do." The guest gives the same answer they've given a hundred times, and the episode sounds like an ad.

The person is more interesting than the company, and the person is who you're building a relationship with. If the show exists to start relationships with buyers or investors, put the spotlight on them as people. The company comes up anyway, later, and it lands better because by then the audience cares about the person behind it.

I told him he didn't have to follow my formula of starting way back. The rule is just that the focus stays on the guest and the spotlight stays on them.

AI drafts, the guest edits

I don't write the questions from scratch. Claude drafts them based on research on the guest: their background, past interviews, what they've written, what their company does. That gets me a strong first draft in minutes instead of an hour.

Then the questions go to the guest ahead of time, in a Google Doc they can edit. Some cross things out. Some rewrite a question to get at what they actually want to talk about. Some add one.

That step does more than prep. It makes the guest a co-author of the episode. They show up knowing what's coming, comfortable, and already invested in it going well. Nervous guests give short answers. Prepared guests tell stories.

Artillery, not a script

"Usually I never get through 21 questions, right? Like, I just have those as artillery."

Twenty-one questions is more than any interview needs. That's the point. You want more ammunition than you'll use, so you're never stuck. If an answer runs long and gets good, you follow it and skip three questions. If a guest gives short answers, you've got plenty left. If a section is landing, you stay in it.

The questions keep you on time, not on rails. A script makes you a reader. A deep bench of questions makes you a host.

How the timing works

I book a 45-minute window for an interview I aim to keep around 20 minutes. The extra time covers the warm-up, tech checks, and the conversation after the recording stops, which is often the most valuable part of the whole thing.

Twenty minutes of recorded conversation is enough for a good episode. The window around it is where the relationship actually starts.

Build yours this week

• Pick three sections in a fixed order that starts with the person.
• Have AI draft seven questions per section from real research on the guest.
• Read every question. Cut anything generic you'd be embarrassed to ask.
• Share the doc with the guest in advance and invite edits.
• Book more time than you'll record.
• Walk in planning to skip half the questions.

The template is the easy part. Booking the right guests, every week, is where shows die. If you want that side running, see how the system works.

Frequently asked questions