How to interview a founder for stories that actually publish

The questions are almost interchangeable. The listening is not.

A mediocre listener with great questions produces a transcript. A great listener with basic questions produces stories. That is the whole job, and it is why most founder interviews come back flat: someone read good questions off a list and waited for their turn to ask the next one. This piece is the opposite of a question list. It is the session mechanics: the setup, the opener, the sequence, the reflexes, and what you do with the recording once the founder logs off. A good session is engineered, not lucky.

This is for whoever is running the extraction. Content operator, strategist, internal comms lead. The mechanics are the same no matter the title.

Set the session up so the recording is usable before the founder says a word

Most of the quality is decided before the call starts. Founders make a thousand decisions a day, so the move is to hand them exactly what they need and nothing more, so they arrive ready to follow your lead instead of treating the session as one more thing to figure out.

Send them four lines, not a brief:

• One 60-minute session, video call, recorded.
• We are capturing raw material for social posts, articles, and clips.
• Wear something you would be fine being clipped on video.
• Prepare nothing. Show up as you are.

Then handle the technology before they connect. A broken call in the first two minutes burns the energy you spent setting it up. Use a tool like Descript that holds quality on a marginal connection and auto-transcribes after, which saves roughly an hour of cleanup on every session. If you run these regularly, set the founder up with a scrappy kit: a desk mic, headphones, a decent webcam. Once tested, it doubles as their podcast and media kit.

Take the first minute to make them look good on camera. Clear shot, decent light, nothing distracting behind them. Ask them to rotate or close a blind if something is off. Fix it in the first minute or you will be distracted by it for the whole hour.

Open in a way that lowers the founder's guard

Start recording the moment they connect, before any "okay, we're starting now."

Two reasons, both about register. The in-between moments (the small talk, the quick aside about something that happened that morning) tend to be the most natural-sounding material you will get all session. And the explicit "we're rolling" announcement pushes founders into media-interview mode, which is the polished, rehearsed register you are trying to avoid.

The session works when the founder forgets the camera is there. You are leading and they are following, so set that frame without apologizing for it: thank them for the time, skip the throat-clearing, move into the work. Even with a founder you find intimidating, the session holds when you hold the structure.

Run a question sequence that is grounded, recent, and specific

Come prepared with questions. Do not feel bound to follow them.

Four openers that consistently produce:

1. What changed in the last ninety days that you did not see coming?
2. Walk me through a sales call from the last month that stuck with you. Any call, good or bad.
3. What decision are you sitting with right now that you keep going back and forth on?
4. What is something a customer said recently that you keep thinking about?

Notice what they have in common. Each one grounds the founder in the last ninety days, where the actual detail lives. Each asks for one specific instance, not a summary. Each opens a door to a moment with stakes inside it.

Now the contrast. The questions that fail are the big thematic ones: "What's your founder story?" "What's your biggest lesson as a CEO?" Those produce rehearsed language, and rehearsed language does not contain stories. It contains the version the founder has told fifty times, sanded smooth, with every interesting edge already filed off.

Your goal is not getting through the list. It is pulling material that is useful for the founder and for the content. Stay on a thread when it is producing. Drop questions that aren't. The prepared sheet is insurance against silence, not a script.

Build the follow-up reflexes that catch the real story

Underneath every good session is one skill: hearing the moment a founder edits a real story out in real time, and stopping them. There are four audible tells:

• The half-second hesitation before they continue in a slightly safer direction.
• The "anyway" pivot that skips past the concrete thing they just started to say.
• The "just" minimizer ("it was just a quick call") flattening a moment they've pre-judged as too small.
• The abandoned specific, a name, date, or number dropped into a summary and left behind.

What each tell sits on top of, and the exact words that send a founder back to the spot they skipped, is the listening discipline covered in full in story extraction. In the session itself, the reflex that matters is the interrupt. You have to break in, and most interviewers are too polite to do it. The trick is interrupting so it reads as interest, not correction. Short. Curious. Name the specific thing.

"Wait, go back to the call with the VP." "Who said that, exactly?" "What did you do the next morning?"

The founder resists for half a sentence, then gives you the real thing. The resistance is just them still believing the specifics are too small. After two or three good interrupts, they start self-correcting, feeling the "anyway" coming and catching themselves. That is the moment the call shifts from interview to extraction.

One more reflex: when something lands well, ask them to say it again and nudge the wording. It feels repetitive; it isn't. Founders want to sound good in the content, so pulling a clean soundbite out of a rough first take is part of the work, not an imposition.

Close on something open. "Is there anything that came up for you that we didn't get to?" It surfaces the thing they were holding, or the idea they had mid-question and never came back to. Some of the best material lands in that last five minutes, which is why you book the full hour even when the substance is 45.

Treat the recording as raw material to extract, not a transcript to file

The work is not done when the call ends. The footage is raw material, and what you do next decides whether the session was worth the founder's hour.

Secure the footage first. Download it, save it, back it up before anything else. If the platform transcribed it, that transcript is now your map.

Then resist the instinct to start clipping. Sit with what was actually said and sort it into four buckets:

• Clippable moments. Short, self-contained, can stand alone as a 30 to 60 second video.
• Quotes. Strong lines worth pulling for post copy or design assets.
• Future material. Half-formed ideas worth coming back to next session.
• Discardable. The rest.

For the clips, cut once, step away, then cut again. A 45-second clip almost always becomes a tighter 25-second clip on the second pass, and the cut almost always comes off the top: the throat-clearing before the founder gets to the point. Punchier openings hold attention. Do not underestimate the second pass. This is the same extraction discipline covered in story extraction, applied to a single recording instead of the whole business.

The Upshot

If you run one founder session and remember one thing, make it this: control the room and spend your attention on the listening, not the question sheet. Set the frame so they follow your lead, start recording cold, ground every question in the last ninety days, and chase the four signals (hesitation, "anyway," "just," the abandoned detail) the second you hear them. The post-call work is extraction, not transcription, and the second pass is where the clip gets sharp.

The sharper point is the one most people get backwards. Founder stories don't sound generic because the founder is boring or because you asked the wrong question. They sound generic because nobody in the room caught the real moment when it surfaced, so the polished version is all that survived to the edit. That's not a question problem. It's a listening problem, and it's the one part of this you can't outsource to a better prompt sheet. For why the generic version is the default, see why founder stories sound generic. For where this sits in the wider practice, the founder communications guide is the pillar.

This is the kind of session that runs underneath DUO's Founder LinkedIn work: one extraction call, then a documented voice profile and AI-assisted drafts the founder reviews, on a budget of two to three hours a month.