← All posts

AI Prompts to Repurpose Content Across Platforms: One Input, Ten Native Outputs

You already made the thing. The blog post is published, the podcast is recorded, the long YouTube video is up. The hard part is supposedly done. So why does it feel like it disappeared into a hole after 48 hours?

Because publishing once is not distribution. The creators who look prolific aren't making ten times more content than you. They're taking one idea and stretching it across every platform their audience lives on, with each version formatted the way that platform actually rewards. One input, ten outputs.

This is different from planning a month of posts. That's a calendar problem. This is a multiplication problem: you have a finished asset in hand and you want to wring every drop out of it. The prompts below do exactly that, and they're written so the output is natively formatted, not the same paragraph pasted into five boxes.

Why "just repost it everywhere" fails

A LinkedIn post that reads like a tweet looks lazy. A tweet that reads like a LinkedIn post is too long and dies in the algorithm. An Instagram carousel needs one idea per slide with a hook on slide one. A YouTube Short needs a spoken script with a pattern interrupt in the first three seconds. A newsletter blurb needs a reason to click, not the whole article dumped in the email.

Each platform has its own physics. The mistake is treating repurposing as copy-paste when it's actually translation. Good AI prompts handle that translation for you, but only if you tell the model the format rules explicitly. Vague prompt in, mush out.

The master repurposing prompt

Start here. This is the one-input-to-ten-outputs engine. Paste your source content (the full blog post, the podcast transcript, or the video transcript) where indicated, and the model returns every format in one pass.

You are my content repurposing engine. I'm going to give you one piece of source content. Your job is to turn it into five platform-native assets. Do NOT copy-paste the same text into each format. Rewrite each one to match how that platform's audience reads and how its algorithm rewards content.

Here is the source content:
"""
[PASTE YOUR BLOG POST / TRANSCRIPT HERE]
"""

First, in 2 sentences, tell me the single core idea worth spreading. Then produce, clearly labeled:

1. X/Twitter thread (6–9 posts). Hook tweet must stop the scroll and make a specific claim, not a vague tease. One idea per tweet. No hashtags. End with a soft CTA.
2. LinkedIn post (180–220 words). First line is a standalone hook with a line break after it. Short paragraphs, plenty of white space, one concrete takeaway, conversational not corporate. One question at the end.
3. Instagram carousel script (6 slides). Slide 1 = hook. Slides 2–5 = one point each, max 25 words per slide. Slide 6 = CTA. Give me the on-slide text plus a one-line caption with 3–5 relevant hashtags.
4. YouTube Short / Reel script (under 45 seconds spoken). Open with a 3-second pattern interrupt. Write it as spoken words with [on-screen text] cues. End on a loop or CTA.
5. Newsletter blurb (120 words). Curiosity-driven, give the reader the "why it matters," and tease — don't dump — the full piece. End with one clear link CTA.

Keep my voice: [direct / warm / no jargon / etc.]. Audience: [who they are].

That single prompt is the whole system. Everything below is for when you want to sharpen one specific output instead of generating all five at once.

Sharpen the X thread

Take this idea: [paste core idea or the relevant section]. Write an X thread of 7 posts. The first post must make a counterintuitive or specific claim that earns the click — no "a thread 🧵" filler. Each following post delivers one standalone insight that could be screenshotted on its own. Keep posts under 240 characters. No hashtags, no emojis except where they genuinely add clarity. Final post: a one-line CTA pointing readers to my free resource.

Sharpen the LinkedIn post

Rewrite this for LinkedIn: [paste section]. Format rules: line 1 is a hook that creates an open loop and stands alone before a line break (because LinkedIn truncates after ~2 lines). Then 5–7 very short paragraphs, one idea each, generous white space. Include one specific example or number. Plain language, zero buzzwords. End with a genuine question that invites comments. 200 words max.

Sharpen the carousel and the Short

These two carry the most "format tax" — they fail hardest when you just paste text. Be explicit about constraints:

Turn this into an Instagram carousel of exactly 7 slides. Slide 1 is a bold hook under 8 words. Slides 2–6 each teach one micro-step in under 20 words with a 2–4 word slide title. Slide 7 is the CTA. Give me each slide's text separately, then write a caption (first line is its own hook) with a 1-sentence summary and 5 niche hashtags my exact audience follows.

Write a 40-second vertical video script from this content. Line 1 must be a spoken pattern interrupt that makes someone stop scrolling ("Stop doing X" / "Nobody tells you that…"). Then 3 fast points, each with an [on-screen text] caption cue. Conversational, like talking to one person. Close with a line that loops back to the hook or sends them to the link in bio.

A repeatable workflow, not a one-off

Once you trust the master prompt, the loop takes ten minutes:

The compounding part: every new piece you publish now becomes five pieces. Do this weekly and you've quietly built a distribution machine without making any more net new ideas. That's the whole point — you're not working harder, you're refusing to let finished work die after one post.

Pro moves most people skip

Feed the model your real voice. Paste two or three of your best-performing past posts and say "match this tone and rhythm." Generic AI output dies on social; your cadence is what makes it land.

Ask for the hook menu first. Before generating full assets, prompt: "Give me 10 hook options for this idea, ranked by scroll-stopping power." Pick the winner, then build around it.

Repurpose across time, not just platforms. A post from six months ago is brand new to 90% of your current audience. Run old winners back through the master prompt with a fresh angle.

Localize the CTA per platform. X gets "free resource in the replies," LinkedIn gets "link in comments," Reels gets "link in bio." Small friction reductions, real difference.

Get the prompts working in five minutes

If you want to test-drive the approach before building your own library, grab the free 10-prompt pack — it includes a starter repurposing prompt you can run on your next post today. No cost, no fluff, just paste-and-go.

And if repurposing is going to be a weekly habit — which is where the real compounding lives — the Content Machine pack is built for exactly this. It's the repurposing-heavy toolkit: the full one-input-to-ten-outputs master prompt plus tuned variants for threads, carousels, Shorts, newsletters, and hooks, all written to produce natively formatted output instead of copy-paste sameness. One finished asset goes in, a week of platform-native content comes out. If you make content and you're tired of it vanishing after a day, that's the engine to plug in.

You already did the hard part by making the thing. These prompts make sure it actually travels.

Get the full toolkit →

FAQ

How many pieces of content can I realistically get from one source? From one solid blog post or video you can repurpose 8 to 12 native pieces — a thread, a LinkedIn post, a carousel, a Short script, a newsletter blurb — without it feeling recycled, as long as you reformat for each platform instead of copy-pasting.

Does Google or social media penalize repurposed content? No. Repurposing is reformatting one idea for different audiences, not duplicating a URL. What platforms flag as spam is posting identical text verbatim across many accounts — varying the format for each platform avoids that entirely.

What's the fastest way to repurpose without sounding robotic? Give the AI the original plus the target platform's format and tone, and ask it to rewrite natively rather than just shorten. Then do a quick human pass so it sounds like you, not a reformatting machine.

Related reading