The short version: one recorded interview, whether you're the host or the guest, can give you ten separate pieces of content if you treat the recording as raw footage rather than a finished thing. The trick isn't clever editing, it's deciding what the ten pieces are before you press record, then mining for them on purpose instead of hoping something usable turns up.
Why most people only get two pieces out of an interview
I did a 40 minute recorded call in January with a fintech founder for a piece I was writing about AI in small business lending. Good conversation, useful stuff about how his team uses AI to flag risky applications before a human even looks at them. I published one article. That was it. For about a week.
Then I sat down and counted what was sitting unused in that one recording. Three strong opinions he'd stated plainly. Two numbers I hadn't used in the article. One moment where he disagreed with something I'd said, on the record, which is gold and I'd just let it sit there. A story about a client who nearly lost his business. That's not one article's worth of material, that's a fortnight's worth.
Most people get two pieces out of an interview, usually the article and maybe a quote graphic, because they edit for the finished piece and throw away everything that didn't fit. The finished piece is the smallest use of the interview, not the main one.
The ten pieces, specifically
Here's the actual breakdown I use now, from that one 40 minute call:
- 1. The main article or blog post - the thing you set out to write, 1200 to 1800 words, built around the interview's central argument.
- 2. A LinkedIn post built on the best single quote - one sentence, the interviewee's name, and 150 words of your own take on why it matters.
- 3. A LinkedIn carousel breaking down one framework they described - the fintech founder described a three step risk check his team runs, that became a five slide carousel with no editing needed beyond formatting.
- 4. A short video clip - 45 to 90 seconds of them saying the most quotable thing, captioned, posted natively rather than as a link out.
- 5. A second short clip on a different point - people assume you can only pull one clip per interview. You can usually pull three or four if you go looking rather than clipping the first good moment you hear.
- 6. An email to your list - written as if you're telling a friend about the conversation, with one line they said that stuck with you.
- 7. A quote graphic - one sentence, their name and title, your branding, posted to Instagram or wherever your audience reads quote graphics (mine barely do, so this one's optional depending on your platform mix).
- 8. A "here's where I disagreed" post - the moment of friction in the conversation, written up as your own opinion piece that references the interview but isn't about them.
- 9. A thread or short-form written breakdown - three to five points from the conversation, numbered, no fluff, works well on X or as a LinkedIn document post.
- 10. A follow-up piece three weeks later - something that happened, a stat that came out, a news story, that lets you say "this reminds me of what [name] told me" and go back to the same well without repeating yourself.
The bit that takes the work: the transcript pass
None of this happens by rewatching the video. You need a proper transcript, and you need to read it like you're mining for sentences, not summarising the conversation. I go through with a highlighter (digitally, obviously) and mark anything that's a strong claim, a number, a disagreement, or a story with a beginning and an end. That's usually 15 to 20 marked moments from a 40 minute chat. Ten pieces of content is a conservative target once you've done that pass, not an ambitious one.
This is where a decent transcription tool earns its keep, and where I'd point you towards a proper look at which content creation tools are worth paying for rather than the free ones that mangle names and numbers. Get the transcript wrong and every downstream piece inherits the mistake, which is the uncomfortable bit nobody mentions when they're telling you how easy repurposing is.
The uncomfortable bit about doing this with AI
Here's what people don't say out loud: if you feed a transcript into ChatGPT and ask for "ten pieces of content from this interview," you'll get ten pieces of content, and eight of them will sound like every other AI-repurposed interview on LinkedIn this month. Flat summaries. Generic takeaways. No personality. The tool isn't broken, you've just asked it the wrong question.
What works better is asking it to help you find the moments, not write the posts. I'll paste the transcript and ask specifically for the three sentences where the person made a claim they didn't fully justify, or the two places where their tone changed, or anything that contradicts common advice on the topic. Then I write the actual post myself, because the voice is the whole point. If you want the longer version of how I use it this way, I've written it up in how to use ChatGPT for content creation without sounding like everyone else. And whatever it drafts for you, check it against the actual recording before you publish, because AI tools paraphrase confidently and wrongly more often than people expect, which is exactly why I always fact check anything AI generated before it goes out under my name.
Spacing it out so it doesn't look like you're milking one conversation
Ten pieces from one interview only works if you don't post them all in the same week. I spread mine across roughly 21 days: the main article and one LinkedIn post in week one, the carousel and a video clip in week two, the email and the disagreement piece in week three, and I keep the follow-up piece in my back pocket for whenever a relevant news story gives me an excuse to return to it. Post all ten in five days and your audience clocks that you're stretching one 40 minute call further than it deserves. Space it and nobody notices, because each piece stands on its own.
If your interview happened to be a podcast appearance rather than something you recorded yourself, the same logic applies but with a different set of angles, and I've laid out the specific breakdown for that in how to turn one podcast appearance into a month of content. The core skill transfers either way: it's about mining, not stretching.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
What this looks like if you hate doing any of this
I know plenty of business owners who'd rather do the actual interview ten times over than sit down and slice one recording into ten pieces. If that's you, the fix isn't forcing yourself through a process you resent, it's building a simpler system: one person or one hour a week dedicated purely to the mining pass, using the transcript and a short checklist rather than starting from scratch each time. I've written more broadly about this for people who find content creation draining rather than just tedious in content marketing for business owners who hate creating content, and the short version is: you don't need to enjoy it, you just need a repeatable checklist so it doesn't take willpower every single time.
Worth saying too, the main article itself isn't the end of the road. Once it's written, it becomes its own source material, and the same mining approach works on it, which is exactly the process I cover in how to repurpose one blog post into a week of content. One interview, one article, one week of further content off the back of the article. That's not ten pieces anymore, that's closer to fifteen or sixteen, and it all traces back to one 40 minute conversation.
The number that matters
Forget the target of ten. The number that matters is how many marked moments you found in the transcript pass. If you only marked four things worth using, you'll get four decent pieces and six weak ones stretched thin, and your audience will feel the thinness even if they can't say why. Fifteen to twenty marked moments from a 40 to 60 minute conversation is realistic if the interview itself was any good. If you're consistently only finding four or five, the problem usually isn't the repurposing process, it's that the questions you asked in the interview were too safe to produce quotable, disagreeable, specific material in the first place.
Frequently asked questions
Do I need the interviewee's permission to repurpose the conversation into ten separate posts?
If it was recorded for public use, a podcast, a webinar, a recorded interview for an article, you generally don't need separate permission for each piece, but it's good practice to tell them roughly how you'll use it and to send them anything that quotes them directly before it goes live, especially the disagreement piece.
How long should the interview be to realistically get ten pieces out of it?
30 minutes is the practical minimum. Under that, you rarely get enough marked moments to fill ten different pieces without repeating yourself. 40 to 60 minutes is the sweet spot, long enough for tangents and stories, short enough that neither of you runs out of energy.
What's the biggest mistake people make when repurposing an interview?
Publishing everything in the same week. It works far better spaced over two to three weeks so each piece reads as its own thing rather than obviously mined from one conversation, and it gives you a reason to talk about the same person or topic again without it feeling repetitive.
Should I use AI to write the ten pieces from the transcript?
Use it to find the moments worth using, not to write the finished post. Ask it to pull out claims, numbers, contradictions and stories from the transcript, then write the actual piece in your own voice and check any facts or figures it summarises before you publish, because AI paraphrases confidently even when it's got the detail slightly wrong.
Related reading: How to Plan a Month of Social Content Without a Big Team and What Is Content Led SEO and How to Build a Strategy Around It.
