How to repurpose a podcast into LinkedIn posts

A podcast episode is the cheapest source of LinkedIn video anybody has, and video is the format that over-indexes hardest in our data: native video was 26.3% of our top-performing decile against 10.1% of the bottom half, a 2.6x gap across 1,613 video posts from 46 creators. To repurpose an episode properly, work from a timestamped transcript and pull three things: one to three clips, three to six claims that stand alone as text posts, and one framework worth a document post. The part that decides whether any of it works is the rewrite around the clip, not the clip.

Why are podcast clips a good fit for LinkedIn?

Two reasons, and only one of them is the data.

The first is the format finding. We scored 12,988 English LinkedIn posts by engagement rate and compared the top decile (1,298 posts) against the bottom half (6,494 posts). Native video appeared more than two and a half times as often among the best posts as the worst, and unlike most format claims on LinkedIn the sample can carry it.

The absolute numbers agree. The median video post in our cohort earned 181 reactions and 27 comments at 0.60 engagement per 1,000 followers, against a cohort median of 122 reactions, 12 comments, and 0.40 per 1,000.

The usual caveats apply and we are not going to bury them. These are correlations, not causes. Our cohort skews toward established creators with 1,000 or more followers across 65 accounts. Engagement counts are lifetime-cumulative at scrape time and post ages vary. And part of what any format measures is the effort it takes to produce, because nobody films, edits, and uploads a video casually. We argue the mechanism at length in our piece on LinkedIn video posts and what 12,988 posts say about them, and there is no reason to re-run it here.

The second reason is the one specific to podcasts, and it is arguably more useful. The effort argument is exactly what normally makes video a bad trade for most professionals: a video costs an hour, a text post costs fifteen minutes, and consistency beats production value. A podcast breaks that trade. The recording already exists. The marginal cost of a 60-second clip is the time it takes to find it and cut it, which is minutes. You get the format that over-indexes without paying the price that usually makes it not worth it.

What can you actually get out of one episode?

From a 45-minute conversation with a guest, here is a realistic inventory. Note that most of it is not video.

That is roughly a month of posting from one recording, at a sane cadence. It is not sixty posts, and anyone promising sixty is counting sentences.

The poll row comes with a caveat we will state once and mean it: our cohort contains only 38 poll posts from 10 accounts. That sample cannot support any conclusion about how polls perform, and we do not draw one. The reason a poll belongs on this list is structural, not statistical: a podcast produces genuine disagreements, and a poll built on a real disagreement is a different object from the “what is your biggest challenge?” polls that fill the feed. Our piece on LinkedIn polls carries the same caveat throughout.

How do you find the clip-worthy moments?

Not by listening back. Listening back is slow, and it selects for moments that sounded good in the room, which is a different thing from moments that work cold in a feed.

Work from a timestamped transcript and read it with four markers:

1. Mark every sentence you would repeat to somebody at a conference. The test is whether you would say it unprompted, without the episode as context. Most 45-minute conversations contain four to eight of these.
2. Mark every number and every specific. Names, dates, amounts, durations. Specifics are what make a clip credible in the absence of everything around it.
3. Mark the hesitations.The places where a guest pauses, backs up, or says “actually, that is not quite right.” Those are usually the moments where somebody stopped performing and started thinking, and they are almost always the best material in the episode.
4. Mark the disagreements. Any point where you and the guest did not land in the same place. A conversation where everybody agrees produces no posts.

What to avoid: picking clips by “best sound bite.” That criterion selects for polish, and polish is the least useful property a feed clip can have. The clips that work are the ones where somebody says a thing they would be mildly nervous about seeing quoted. The smooth ones are the ones people scroll past.

One more filter worth applying: does the clip survive losing the question? If the moment only makes sense as an answer to something the host asked, it either needs the question written into the caption or it is not a clip.

Why does a clip need a rewrite before it works?

Because the recording was made for an audience that chose to be there, and the feed did not choose. A clip that starts mid-sentence in a conversation nobody heard the beginning of is the single most common bad podcast post on LinkedIn, and it is bad for a structural reason rather than a quality one. The listener had twenty minutes of context. The scroller has none.

Three fixes, in order of how much work they cost.

Cut so the first sentence is a complete claim. The cheapest option and usually the right one. Find the point in the transcript where the speaker states the thing rather than building toward it, and start there. Everything before that is setup the feed will not sit through.

Write the context into the caption. The caption is not a summary of the clip. It is the thing that replaces the twenty minutes the viewer missed. One line naming who is talking and what question they are answering does most of the work.

Record a five-second cold open.More work, and it is the version that actually looks intentional. You, to camera, saying the setup in one sentence, then cut to the clip. This also solves the problem where a guest's clip has no visual identity connected to your account.

The general point: the caption carries the context the listener would have had.Everything else about clip production is secondary to that. A well-edited clip with a caption that says “great conversation with X, full episode below” will underperform a rough clip with a caption that states what the clip argues.

How long should a podcast clip be on LinkedIn?

Thirty to ninety seconds covers almost everything. Native in-feed video runs from three seconds to ten minutes per Sprout Social's social video specs, but the ceiling is not a target, and our general length guidance lives in the video posts piece.

Two cutting rules are specific to podcast material and worth stating separately.

End on the claim, not on the end of the exchange.Conversations trail off. The speaker finishes the point, then adds a qualifier, then the host says “interesting.” All of that is real conversation and none of it belongs in the clip. Cut on the last word of the actual claim, even if it feels abrupt. Abrupt is fine. Trailing off is not.

Leave out the host's follow-up question. It reads as a teaser for the episode, which turns the clip into an advertisement. If the follow-up genuinely improves the clip, that is a sign you cut in the wrong place.

And a length heuristic that is more useful than a number: if you cannot state what the clip argues in one sentence, it is not one clip. It is two, or it is nothing.

Do podcast clips need captions?

Yes, and podcast clips have less excuse than any other video format. The audio is the content. A clip of two people talking, playing muted with no captions, is a picture of somebody's face. Video autoplays muted in the feed, so that is the default state for most of the people who see it.

The accessibility case is the more important one and gets mentioned less. An uncaptioned clip excludes deaf and hard-of-hearing viewers completely, along with anybody in an open-plan office, on a train, or reading in a second language.

The practical point specific to podcasts: you already have the transcript. Generating a caption file from material you already produced costs nearly nothing, which removes the one excuse people usually offer. Burn captions into the clip so they survive being downloaded and reshared, and upload a caption file as well so the text is machine readable. Then read them before publishing, because automatic transcription is reliable until it meets a product name or an accent and then it is confidently wrong.

How do you turn the transcript into text posts?

This is where most of the value is, and almost every podcast repurposing guide skips it in favour of clips. A transcript is a document of somebody thinking out loud for 45 minutes. That is a richer source than most blog posts.

The method:

1. Pull the claim, drop the scaffolding.Spoken language is full of “I think,” “you know,” “kind of,” and half-finished restarts. Strip all of it. What remains is usually one sentence, and it is usually sharper than the original.
2. Fix the person.A transcript is full of “we” addressed to a co-host. A post is “I” addressed to one reader. In our cohort, first-person openers ran 19.6% among top-decile posts against 10.1% in the bottom half, and “you” appeared in 47.5% of top-decile hooks against 32.7% of the bottom half.
3. Attribute anything that was the guest's.By name, in the post, not in a tag at the bottom. Passing off a guest's insight as your own observation is the fastest way to burn the relationship that made the episode possible.
4. Add what you thought about it. A restated claim is a summary. A restated claim plus your reaction is a post. This is the step that separates repurposing from transcription.
5. Put the claim inside the fold. The visible portion runs to roughly 210 characters on desktop and about 140 on mobile, so the claim cannot wait for paragraph two.

For the story-shaped material, which podcasts produce more of than any other source, the craft is different again: one scene, present tense, no framing. Our pieces on how to open a LinkedIn post and telling a story in a LinkedIn post cover the mechanics.

How do you use quote posts without them looking lazy?

A pull-quote on its own is a repost of somebody else's sentence with a nicer typeface. It gives a reader nothing to respond to, and it says nothing about you except that you were in the room.

The version that works has three parts:

• The quote, attributed clearly.Name the person and the context. Not “a guest on my podcast said.”
• Whether you agree, and why.Agreement is fine. Vague agreement is not. “This is why we stopped doing X” is a position. “So true” is not.
• What you did about it, or will. The concrete consequence. This is the part that makes the post yours rather than theirs.

There is now a platform reason to care as well. In March 2026, LinkedIn VP of Engineering Tim Jurka said, in a statement reported by Social Media Today, that LinkedIn would be “improving our systems to reduce repetitive, click-driven posts and filter out engagement bait, so your Feed feels more relevant to your interests, and not a popularity contest.” A bare pull-quote with no perspective attached is a reasonable description of a repetitive post, and it was already a weak one.

Should you tag the guest, and how?

Yes, when the mention is real. Tagging the guest puts the post in front of their network, and a guest who reshares it is worth considerably more than the tag itself, because a reshare comes with an implicit endorsement.

Three rules that keep this from going wrong:

• Tag people who are actually in the post. Not their company, not their three colleagues, not the four other podcasts they have been on. Tagging people who have no connection to the content is the oldest reach trick on LinkedIn and it reads as exactly that.
• Tell them before you post. Send the clip. Guests who see a clip in advance reshare it far more often than guests who discover it in their notifications, and it gives them a chance to flag anything they were misquoted on.
• Make the post worth resharing.A guest will reshare a post that makes them look thoughtful. They will not reshare “great chat with X, listen here.”

Do not organise reciprocal engagement around this. Arranging for a group of people to reliably like each other's posts is a different activity with a different risk profile, which we covered in the piece on engagement pods.

What does a month of posts from one episode look like?

Here is a schedule built from a single 45-minute episode, assuming four posts a week and that the episode is one of several inputs rather than the only one.

Note what is missing: there is no “new episode out now” post. If you want one, make it one post, early, and accept that it will be the weakest performer of the eight. Slot the rest into whatever cadence you already sustain rather than adding to it. Our pieces on how often to post and building a content calendar cover how to fit a source like this into a schedule without letting it take over.

What should you not repurpose from a podcast?

• The full episode. Uploading 45 minutes to the feed asks a scrolling reader for a commitment they did not agree to. Nobody watches it, and the completion numbers you get back are demoralising and meaningless.
• The trailer. A trailer is an advertisement for content. A feed post has to be the content.
• The episode announcement with a link. This turns the post into a link share, and link posts were 22.1% of our top decile against 32.7% of our bottom half across 3,813 posts from 60 accounts. Put the clip up natively and mention the episode in a comment. Our piece on external links covers the evidence and the parts of it that are genuinely contested.
• Any clip that needs the previous ten minutes. If you find yourself writing a three-line caption to explain the setup, the clip is the wrong clip.
• Audiograms with a waveform and no faces. They were a workaround for platforms that did not support what LinkedIn supports. A static waveform is a video that does not move.

How do you measure whether the clips are working?

Impressions are the least useful number here. LinkedIn reports video-specific metrics that no other format gives you, and they answer better questions.

• Completion rate. The most honest number available. If it collapses in the first ten seconds, your cut point is wrong, not your topic.
• Comments over reactions. Comments are the scarcer signal in our data by a wide margin: the top-decile median was 72 comments against 5 in the bottom half, compared with 235 reactions against 80. Video posts in our cohort had a median of 27 comments against 12 cohort-wide.
• Whether the guest reshared. A binary, and a good one.
• Whether anyone listened. If your goal is downloads, the honest measure is downloads, and a clip that performs beautifully while sending nobody to the episode is telling you something about what the post was actually for.

For a baseline, our whole-cohort benchmarks are a median of 0.40 engagement per 1,000 followers, 1.27 at the 75th percentile, and 5.95 at the top-decile threshold. The percentile tables live in our benchmark piece on what good looks like at your follower count. Compare clips against your own median rather than against anyone's benchmark, including ours.

What is the short version of repurposing a podcast to LinkedIn?

Work from the transcript, not the audio. Cut one to three clips so each starts on a complete claim and ends on the claim rather than the conversation, and write a caption that supplies the twenty minutes of context the scroller does not have. Then mine the transcript for the three to six claims that stand alone as text posts, because that is where most of the volume comes from. Caption every clip. Space it all across a month.

The reason to bother is the number at the top: native video was 26.3% of our top-performing decile against 10.1% of the bottom half, across 1,613 posts from 46 creators, and a recorded episode is the only situation in which that format costs almost nothing to produce. The wider workflow across every source you have is in our LinkedIn content repurposing workflow, and the written-source version is in how to repurpose blog posts into LinkedIn posts.