Off Prompt

AI Tools for Small Business

Marketing

Using AI to create a simple content repurposing system that turns one blog post into a week of social content

Repurpose one blog post into social media content with AI using one prompt. Get 7 ready-to-post assets for LinkedIn, Instagram, Facebook and X in under an hour.

Owen Grant 8 min read
Using AI to create a simple content repurposing system that turns one blog post into a week of social content

You write a solid blog post, hit publish, share it once on LinkedIn, and then... nothing. Back to staring at a blank caption box three days later wondering what to post. This guide shows you how to repurpose one blog post into a week of social media content with AI — LinkedIn, Instagram, Facebook, X — using a free AI tool and one well-built prompt. You don't need to be a content strategist or a social media manager to pull this off.

What you need before you start

ChatGPT{:target="_blank"} — OpenAI's AI tool that can read your blog post and rewrite it for different platforms. The free version (GPT-4o mini) handles posts up to around 1,200 words just fine. If your posts run longer, the paid tier ($20/month) removes that limit. Alternatively, Claude{:target="_blank"} (free tier available) has a larger reading capacity and is excellent if you write longer posts or want to paste several at once.

Buffer — a free social scheduling tool you'll use to organize and post your content. Buffer's free plan{:target="_blank"} covers up to three social channels, which is plenty to start.

Time required: About 45–60 minutes the first time, closer to 30 minutes once you've done it twice and have your prompt saved.

Skill level: If you can copy and paste, you can do this.


What "a week of social content" actually looks like — the 7-asset breakdown

Before you touch the AI, it helps to know what you're building. A single 1,000–1,500 word blog post has enough raw material for:

  1. A LinkedIn long-form post (your main piece — 1,200–1,500 characters)
  2. Three short LinkedIn or Facebook posts (one punchy insight per post)
  3. An Instagram carousel outline (5–7 slides, 15–25 words each)
  4. An X/Twitter thread (5–7 tweets, each under 220 characters)
  5. A quote graphic caption
  6. An email newsletter intro paragraph
  7. A short-form video script (45–60 seconds, talking-head style)

That's seven assets from one piece of writing. You won't always use all seven — pick the platforms that actually matter to your business. A plumbing contractor might focus on Facebook and email. A freelance designer might lean into Instagram and LinkedIn. The point is you have options, and the AI does the heavy lifting to get you there.


How to repurpose blog post content into social media posts with AI: the master prompt

This is the core of the whole system. One prompt, sent once, returns all your assets formatted and ready to edit.

Step 1. Open ChatGPT or Claude in your browser and start a new chat.

Step 2. Copy your entire blog post and paste it into a text document for a moment — you'll need it shortly.

Step 3. Type or paste this master prompt first, before pasting your blog post. The reason the prompt comes first is that you're setting the instructions before handing over the raw material — like briefing a copywriter before handing them your notes.

Here's the prompt to copy. This version is intentionally specific about how to find the angles — not just "summarise this," which produces flat, lifeless content, but "find the most useful tip, the sharpest claim, and the most relatable frustration" — which is where real engagement comes from.

You are a social media content writer for a small business owner. I'm going to paste a blog post below. Using only the content in that post, create the following 7 social media assets. Do NOT summarise the post — instead, find the most useful practical tip, the most provocative or counterintuitive claim, and the most relatable frustration the post addresses. Write from those angles.

Format each asset clearly with a bold header.

1. LinkedIn Long-Form Post — 1,200–1,500 characters. Start with a single-line hook. Use short paragraphs and line breaks. End with a question to drive comments. 2. LinkedIn Short Post A — 300–400 characters. One insight from the post. No fluff. 3. LinkedIn Short Post B — 300–400 characters. A different angle. Could be a "hot take" or a practical tip. 4. Facebook Post — 2–3 sentences + a question at the end. Conversational tone. 5. Instagram Carousel Outline — Slide 1: hook (10–15 words). Slides 2–6: one idea per slide, 15–25 words each. Slide 7: call to action. 6. X/Twitter Thread — 5 tweets, each under 220 characters. Tweet 1 is the hook. Number them 1/ through 5/. 7. Email Newsletter Intro — 3–4 sentences that tease the blog post and include a "read more" placeholder link.

Here is the blog post: [PASTE YOUR BLOG POST HERE]

Step 4. Replace [PASTE YOUR BLOG POST HERE] with your actual blog post text and send the message.

Step 5. Wait 30–60 seconds. You'll get all seven assets back in one response, clearly labelled.

The output won't be perfect — and that's fine. You're not replacing your voice. You're getting a first draft of seven things instead of starting from zero on seven things. That's the real time saving.


Platform-by-platform formatting: what to tweak before you post

The AI gets you close, but each platform has quirks worth knowing.

LinkedIn rewards posts that start with a single bold line before you hit "see more." Make sure your hook is on its own line. Add line breaks between paragraphs — a wall of text gets skipped.

Instagram carousels need a visual in your mind as you edit. Each slide is basically a headline. If a slide runs over 25 words, cut it. The goal is one idea per swipe.

X/Twitter threads live or die by Tweet 1. If the hook doesn't stop the scroll, nobody reads Tweet 2. Read your first tweet out loud — if it sounds like a press release, rewrite it.

Facebook is the most forgiving. Posts that ask genuine questions (not "agree?" — something with a real answer) tend to pull more engagement than statements.

Spend 10–15 minutes here. Small tweaks, not rewrites.


How to build your brand voice into every AI output (so it sounds like you)

Here's the simplest fix for AI content that sounds generic: add a voice paragraph to the top of every prompt.

Before the master prompt above, paste a short description of your tone and your business. You only need to write this once, then save it somewhere you can copy it quickly.

My business is [what you do] for [who you serve]. My tone is [describe it — warm and direct? Informal and funny? Professional but plain?]. I avoid [any phrases or topics that feel off-brand]. My audience cares most about [main concerns or goals they have].

That's it. Three sentences. The AI uses it to calibrate the language, word choice, and energy of everything it writes for you. A salon owner might write "friendly and confident, not clinical." A contractor might write "no-nonsense, practical, working-class — no marketing-speak." HubSpot's 2024 State of Marketing report{:target="_blank"} found that marketers who repurpose content consistently report roughly three times the output without a matching jump in time — and a brand voice prompt is what keeps repurposed content from feeling like it came from a machine.


Putting it all together: a simple weekly workflow that takes under an hour

This is what the whole thing looks like when it's running smoothly.

  1. Write your blog post (or use one you already published — yes, old posts work fine).
  2. Open ChatGPT or Claude. Paste your voice paragraph, then the master prompt, then your post. Send.
  3. Copy each asset into a document or directly into Buffer{:target="_blank"} or Meta Business Suite{:target="_blank"} (free for Facebook and Instagram scheduling).
  4. Edit for 20–30 minutes. Add a local reference, swap a word that doesn't sound like you, make a tip more specific to your clients.
  5. Schedule the posts across the week — not all at once. Spread them out so each one gets a fair shot.

Total time: 30–60 minutes. Compare that to the 3–4 hours most small business owners spend creating social content from scratch each week, and you'll see why this workflow is worth the one-time effort to set up.


When AI repurposing doesn't work well — and what to do instead

The posts sound flat and generic. This happens when the prompt says "summarise" instead of "find the best angle." Go back and use the master prompt above as written — the specific instructions about provocative claims and relatable frustrations make a real difference.

The AI misses your specific examples or local details. AI works from the words on the page. If your blog post doesn't mention your city, your clients' specific situations, or a real story from your business, the output won't either. Fix: edit those details back in during your 20–30 minute review. This is normal. It's a first draft, not a finished post.

The output is too long for the platform. Most AI tools default to "more is more." If a LinkedIn short post comes back at 600 characters instead of 350, just add "Keep each short post under 350 characters" to the prompt. Constraints help.


What to do next

Save your master prompt and your voice paragraph somewhere easy to find — a Google Doc, a sticky note in Notion, anywhere you'll actually remember to grab it. The second time you run this workflow, it'll feel like a system instead of a project. If you want to extend this further into email marketing, there's a solid walkthrough on turning AI drafts into a real email newsletter workflow coming soon on this blog.


FAQ

Can I use an old blog post, or does it have to be brand new? Absolutely use old posts — some of your best evergreen content is probably sitting there unread. Pull something from six months ago that your current audience hasn't seen. The repurposing process works exactly the same way.

What if I don't have a blog at all? You can use any long-form content: a detailed email you sent a client, a proposal you wrote, a FAQ document, even a voice memo transcript. As long as there's 500+ words of useful content, the AI has enough to work with.

Is the free version of ChatGPT good enough, or do I need to pay? The free tier handles posts up to around 1,200 words without issue. If your posts run longer, paste them in two parts, or try the free version of Claude{:target="_blank"}, which has a much larger reading capacity at no cost.

Do I need a separate tool for scheduling, or can I just post manually? Manual posting works fine if you're disciplined. But scheduling tools like Buffer or Meta Business Suite remove the friction of remembering to post — you do it all in one sitting, then forget about it for the week. Worth the five minutes of setup.

Will people know the content was written by AI? Not if you edit it. Generic-sounding posts happen when people skip the editing step — the 20–30 minutes of review is what makes it sound like you. Think of the AI output as a rough draft, not a finished product. You're the editor; the AI is the intern who did the first pass.

Was this useful? ·