Blog • Content

How to Turn a Podcast Into LinkedIn Posts

A prompt that turns a podcast excerpt into three LinkedIn posts without adding anything the speaker didn't say, tested on a short transcript.

By Akesh Chandrasiri20 October 20263 min read

A forty-minute podcast contains a month of LinkedIn posts. The hard part is picking the moments worth posting and keeping them true to what the speaker said. AI can help with the first and needs a firm leash for the second.

Start with permission

If the podcast isn't yours, ask. For a client, get written agreement on what you may quote and how it will be credited. For someone else's show, don't repurpose their words without permission. Quote briefly, credit the speaker and link to the episode.

Step 1: Find the moments

Read or skim the transcript and mark the passages where the speaker:

  • Says something surprising.
  • Describes a mistake or a change of mind.
  • Gives a specific method or a rule of thumb.

One strong passage of five to ten sentences is enough for several posts.

Step 2: Use a strict prompt

Turn this podcast excerpt into 3 LinkedIn posts, each under 110 words, each with a different angle. Use only what the speaker says; do not add statistics, results or stories. Keep the speaker's phrasing where it works. Start each post with a first line that stands alone. No hashtags, no emojis, no questions that beg for comments.

EXCERPT: [paste]

The key rule is "use only what the speaker says". Without it, the AI adds background and conclusions the speaker never gave.

The test

I wrote a short invented excerpt about moving from hourly pricing to project pricing:

I used to charge by the hour, and it punished me for getting faster. The better I got, the less I earned. So now I quote per project. I work out roughly how long it will take, add a buffer, and I tell the client the number up front. The scary part is the first time you say a number out loud. But the clients who push back hardest on the price are usually the ones who'd have argued about the hours anyway.

Claude's three LinkedIn posts from the excerpt

The results were three posts with three angles: the incentive problem with hourly rates, the fear of saying the first number, and who pushes back on price. Each opened with a line from the excerpt that could stand alone. They kept the speaker's phrasing, such as "the better I got, the less I earned".

The part I liked best

For the third post, the AI added a closing line ("If someone is going to dispute the cost, I'd rather it happen before the work starts...") and then flagged it itself. It said that line was its inference from the speaker's point, not something the speaker said, and suggested cutting it to stay strictly faithful.

That is the behaviour to look for, and to ask for. Always read the notes the AI adds after the posts. I would cut that line, because putting words in someone's mouth is how repurposed content goes wrong.

Step 3: Check against the transcript

Read each post next to the excerpt and ask:

  • Did the speaker actually say this?
  • Is a number, name or result in here that isn't in the transcript?
  • Has the meaning shifted? "Usually" can drift into "always".

Step 4: Add the credit

Add the speaker's name and the show at the end of each post, with a link to the episode.

The takeaway

Repurposing a podcast works when you treat the transcript as the only source of truth. Ask for posts that use only what was said, read the AI's own notes, and check every line before it goes out under someone's name.