I pulled 14 vertical clips out of one 62 minute interview last month. Nine of them came from four questions. The other 48 minutes of that conversation, all of it perfectly pleasant, produced almost nothing I could cut.
That ratio is not a talent problem. The guest was sharp. The audio was clean. The host was warm and prepared. The problem was that most of the questions asked for a summary, and a summary is the one thing you cannot clip.
I edit podcasts for a living, which means I watch the whole episode. Every minute of it, with a human brain, not a silence detector. So I have sat through a lot of interviews looking for the moment where a person stops performing and starts saying something. After a few hundred of those, the pattern is boring and obvious: clippable answers come from questions that force a specific memory, not a general opinion.
The summary trap
Here is the question I see most often in the first five minutes of an interview: "So tell us a bit about your journey."
I understand why it is there. It is polite. It hands control to the guest. It fills space while both people settle in. And it produces four minutes of resume that I will trim down to fifteen seconds if I am lucky.
The reason is mechanical. When you ask someone to summarize, they compress. They flatten ten years into a list of job titles and lessons. Compression removes exactly the things that make a clip work: a scene, a number, a stake, a moment where the outcome was uncertain.
You cannot edit compression back into detail. I can cut a long answer down. I cannot cut a vague answer up.
Ask for a scene, not a stance
The single change that moves the most: replace "what do you think about X" with "when did X actually happen to you".
Watch the difference in what comes back.
- "What do you think makes a good hire?" gives you three adjectives and a maxim.
- "Tell me about the last person you hired who did not work out." gives you a room, a date, a decision, and usually a flinch.
The second answer is clippable because it has a shape. Something happened, then something changed. That is the smallest complete story, and it is the only structure that survives being pulled out of context and dropped on a phone screen with no setup.
When I go looking for material for clips, I am scanning for that shape. Not for quotable wisdom. Wisdom without a scene attached reads like a poster. People scroll past posters.
Numbers do half the work
Ask for a number and you get a hook for free.
"How long did that take?" "What did it cost you?" "How many did you send before the first reply?" These questions feel almost rude in their smallness, which is exactly why they work. A guest cannot generalize their way past a number. They either know it or they say they do not, and both answers are more interesting than an abstraction.
The most reliable clip opener I see is a guest saying a specific number in the first three seconds. Not a claim, a number. It buys you eight seconds of attention that no amount of clever editing can manufacture.
This is also why finance and technical guests clip so well when they are asked properly. The specificity is already in their head. Someone has to reach in and ask for it. We do a lot of podcast editing for accountants and the shows that perform are never the ones explaining what a balance sheet is. They are the ones where the host asks what the worst client mistake of the year actually cost, and the guest answers with a figure.
The follow up is the question
Most of the good material in any episode I cut arrives on the second or third pass at the same subject, not the first.
The first answer is the rehearsed one. Everybody has a version of their story they have told forty times. It is smooth, it is safe, and it is useless to me. The second answer, the one that comes after "wait, go back, what happened right before that", is where the flinch lives.
Three follow ups that earn their keep:
- "What did that feel like at the time?" Pulls the guest out of narration and back into the moment.
- "Who told you that?" Turns an abstract lesson into a person in a room.
- "What would you have done instead?" Forces a concrete alternative, which is where opinions get sharp.
The mistake is treating your prepared list as a checklist. If you are looking at question seven while the guest is still answering question six, you will miss the sentence I was going to build the whole clip around. I have watched this happen on the timeline more times than I can count. There is a small pause, a hint of something real, and then the host says "great, so moving on".
If a guest is giving you short answers and you are not sure whether to push, that is a different problem with a different fix, and I wrote about it in what to do with a quiet guest.
Questions that are secretly about you
There is a category of question that sounds generous but is really the host talking. You know the version: a ninety second setup, three clauses of context, a personal anecdote, and then "so, thoughts?"
I have to cut those. Not because the host is boring, but because a clip cannot carry two people's context. The vertical format gives you maybe two seconds before someone decides. If the first two seconds are the host clearing their throat, the clip is dead.
Short questions produce long answers. Long questions produce short ones. This is close to a law.
The exception is a co-hosted show, where the back and forth is the product and a bit of overlap reads as chemistry rather than clutter. That only works when the chemistry is genuinely there, which is not something I can build in post. I have been blunt about that in co-host chemistry cannot be edited in.
Build the question around the guest, not the format
The best interviewers I cut for do something that looks like extra work and is actually a shortcut. They find the one thing this guest knows that nobody else on their booking list knows, and they spend twenty minutes there instead of covering nine topics at five minutes each.
Nine shallow topics give me nine unusable segments. One deep topic gives me a clip, a chapter, and usually the episode title.
This matters more when you are early and your guests are not famous. If the name is not the draw, the specificity has to be. A guest with 800 followers and one genuinely strange story will out perform a well known name giving their standard talk, every single time. That is most of the argument in booking guests without a big audience, and it holds up in my timeline.
Musicians are a good example of this. Ask a working musician about "the creative process" and you get fog. Ask what they cut from the second verse and why, and you get craft, and often a demonstration. Half the reason podcast editing for musicians is fun is that the good answers come with something to show.
What this looks like in the edit
A practical note on how this lands on my end, since the whole point of a clippable answer is that somebody has to find it.
We deliver a first cut in 48 hours, and revisions come back the next working day. Vertical clips are $25 each as an add-on, and show notes with chapters are $35. None of that changes based on how good your questions were. What changes is how many clips are actually worth making.
On a well interviewed 60 minute episode I can usually pull eight to twelve clips that stand on their own. On a badly interviewed one of the same length, I will find three, and two of them will be the host reading the guest's bio back to them in a slightly more interesting way. Same edit time, same price, a third of the output.
So the highest leverage thing you can do for your own output math is not to buy more add-ons. It is to ask four better questions.
A few small mechanical things that help me find the moment faster:
- Leave a beat of silence after a big answer. Do not step on the last three words. I need the tail to make a clean out point.
- If the guest says something great and then talks over it, ask them to say it again. Nobody minds. It reads as care.
- Say the guest's answer back in your own words occasionally. It gives me a second, cleaner version of the same idea to choose from.
- Do not fix your own question mid sentence. Stop, breathe, ask it cleanly. Two seconds of your patience saves a messy cut.
And when a clip does land, the packaging carries it the rest of the way. If it is going on YouTube, the thumbnail is doing more work than anything I did in the timeline, which is why my sister studio ClickTheory exists.
The three questions I would keep
If someone made me throw out every interview prompt except three, I would keep these.
"When was the last time you were wrong about this?" It is disarming, it is specific, and almost nobody has a rehearsed answer for it.
"What does that cost?" Money, hours, relationships, whatever fits. Cost forces a number and a stake in one move.
"What happened next?" The laziest great question in the world. Ask it four times in a row and you will get a story out of almost anyone.
None of this requires a better microphone or a bigger guest. It requires deciding, before you hit record, that you would rather have one real answer than nine polite ones.
Ask the small question. Wait longer than feels comfortable. The rest of it is my job, and I would much rather spend that job cutting down something good than hunting through an hour for something usable. If you want a look at what we do with the recording once you send it over, how it works covers the whole path from upload to final file.