Subscribe →
Video to Text 29 May 2026

How to Turn 1 YouTube Video Into 10 LinkedIn Posts Using Claude (Step-by-Step AI Workflow)

Time saved: ~3–4 hours per video. Cost: $0 if you’re on Claude’s free tier.

You filmed a 20-minute YouTube video. You edited it, uploaded it, wrote the description — and then you posted it and moved on. Meanwhile, that single video contains enough insight to fuel a week’s worth of LinkedIn content. Most creators never extract it.

This workflow fixes that. In under 30 minutes, you’ll have 10 ready-to-post LinkedIn posts pulled directly from one video. No paraphrasing. No staring at a blank screen. Just a proven Claude prompt, a transcript, and a repeatable system.

The Stack: What You Need

  • YouTube video (yours or one you have permission to repurpose)
  • Claude (Sonnet 4 or higher — the free tier works)
  • Transcript source — one of the following:
    • YouTube’s built-in transcript (free, under the ... menu on any video)
    • Tactiq Chrome extension (free tier, pulls transcripts instantly)
    • Otter.ai or Riverside.fm if you record your own content
  • LinkedIn account (obviously)
  • Optional: Notion, Google Docs, or Typefully to store and schedule your posts

Step 1: Pull the Transcript

Using YouTube’s Built-In Transcript (Fastest, Free)

  1. Open your video on YouTube.
  2. Click the ... (three dots) below the video title.
  3. Select “Show transcript.”
  4. A panel opens on the right. Click the three-dot menu inside the panel and select “Toggle timestamps” to turn them off — this gives you cleaner text.
  5. Select all the transcript text, copy it.

Using Tactiq (Recommended for Longer Videos)

  1. Install the Tactiq Chrome extension.
  2. Open your YouTube video.
  3. Click the Tactiq icon in your browser — it auto-generates and cleans the transcript.
  4. Copy the full text.

Pro tip: Don’t worry about a messy transcript. Claude handles filler words, incomplete sentences, and run-ons just fine. You don’t need to clean it up.

Step 2: Open Claude and Set the Stage

  1. Go to claude.ai and start a new conversation.
  2. Paste your transcript into the chat — just drop it in as plain text.
  3. Add a single line before the transcript: Here is the transcript of my YouTube video:
  4. Hit send. Wait for Claude to confirm it received the content. (It will usually summarise it briefly or say it’s ready.)

Now you’re set up for the main prompt.

Step 3: Run the Master Prompt

Copy the prompt below exactly. Paste it as your next message in the same conversation.

The Master Prompt

You are a LinkedIn content strategist helping a [your niche, e.g. "solo consultant", "B2B SaaS founder", "personal finance creator"] grow their audience with high-performing posts.

Using ONLY the transcript I provided above, generate 10 LinkedIn posts. 

Follow these rules for every post:

1. FORMAT: Each post must be optimised for LinkedIn — short punchy opening line (the hook), line breaks every 1–2 sentences, and end with either a question to drive comments OR a single clear call to action.

2. VARIETY: Use a mix of these post types across the 10 posts — do not repeat the same type more than twice:
   - Contrarian take (challenge a common belief from the video)
   - How-to / tactical tip (one specific actionable step)
   - Personal story or anecdote (pulled from anything personal mentioned in the video)
   - Listicle (3–5 bullet insights, formatted as a clean list)
   - Single bold statement + explanation (one sentence claim followed by a short paragraph)
   - Before/after or problem/solution framing
   - Quote-style post (pull a strong quote or key phrase directly from the transcript and expand on it)

3. TONE: Match the tone of the original speaker in the transcript. If they are casual and direct, keep it casual. If they are professional, keep it professional. Do not make the posts sound like generic AI content.

4. LENGTH: Each post should be between 100–250 words. No longer.

5. HASHTAGS: Add 3–5 relevant hashtags at the end of each post.

6. OUTPUT FORMAT: Number each post 1–10. Add a label for the post type used (e.g. [Contrarian Take], [Listicle], etc.) above each post so I can see what format was used.

Begin.

Step 4: Review and Refine

Claude will generate all 10 posts in one shot. Here’s how to quality-check them:

Check for these green flags:

  • The hook (first line) stops the scroll — it’s specific, not vague.
  • The post sounds like a human wrote it, not a press release.
  • Each post stands alone — a reader doesn’t need to have seen the video to understand it.

If a post misses the mark, use these follow-up prompts:

Post 3 is too generic. Rewrite it using a more specific example or data point from the transcript.

Rewrite post 7 as if the author is speaking directly to a [target audience, e.g. "freelance designer"] who is struggling with [specific problem].

Give me 3 alternative hooks for post 5.

Claude keeps the full context of your conversation, so you can iterate on individual posts without re-pasting the transcript.

Step 5: Store and Schedule Your Posts

Once you’re happy with the 10 posts:

  1. Copy all 10 into a Google Doc or Notion page. Create a simple table: Post number | Post type | Text | Scheduled date | Status.
  2. Space them out. 10 posts from one video = roughly 2 posts per week for 5 weeks. That’s over a month of LinkedIn content from a single recording session.
  3. Optional — use a scheduler: Taplio, Typefully, or Buffer all let you queue LinkedIn posts in advance. Paste your posts in, set your schedule, and walk away.

Advanced Variations: Getting More Out of This Workflow

Turn the Posts Into a Thread or Newsletter Section

After generating your 10 posts, run this prompt:

Using the same transcript, write a LinkedIn carousel outline — 7 slides. Slide 1 is the hook/title, slides 2–6 are the key points, slide 7 is the CTA. Keep each slide to 1 sentence + 1 supporting line.

Batch Multiple Videos at Once

If you have 3–4 older videos sitting in your archive, repeat this workflow for each one in separate Claude conversations. You can build a content bank of 30–40 posts in an afternoon.

Repurpose the Output to Other Platforms

Once you have your LinkedIn posts, run this quick follow-up:

Take post numbers 2, 5, and 8. Rewrite each one for Twitter/X — max 280 characters, punchy, no hashtags.

Expected Output: What “Done” Looks Like

When this workflow is complete, you should have:

  • 10 numbered LinkedIn posts in Claude’s chat window, each labelled with its post type.
  • A variety of formats — at least one listicle, one contrarian take, one how-to, and one story-driven post.
  • Posts that are ready to copy-paste with only minor personalisation needed (adding your name, swapping in a specific client example, etc.).
  • Each post has 3–5 hashtags and ends with a hook question or CTA.

If your output looks like 10 variations of the same bland summary post — that’s a sign the transcript was very thin or the video lacked concrete insights. In that case, go back and prompt Claude to “be more specific and pull exact examples, phrases, and moments from the transcript rather than summarising.”

Why This Workflow Actually Works

Most AI content prompts fail because they’re too vague. They ask Claude to “write LinkedIn posts about [topic]” with zero source material — so it defaults to generic advice that sounds like every other AI post on the feed.

This workflow is different because Claude is constrained to your actual words and ideas. The output reflects your thinking, your examples, your tone. That’s what makes it publishable without a major rewrite.

One video. Thirty minutes. Ten posts. That’s the math.

Found this useful? The next workflow in this series covers how to turn a single blog post into a 5-part email sequence using Claude — same approach, different channel.