Draft a Month of Social Posts from One Article

Workflow area: Marketing · Tier: Foundation · Hands-on time: ~30 minutes

The mistake is asking for "20 social posts from this article" in one go — you get twenty rephrasings of the same summary, each with the same AI-flavored enthusiasm. The recipe below works like an editor: extract the angles first (claims, numbers, contrarian takes, questions), then draft posts per angle per platform, then pass everything through your voice filter. Same effort, but the output reads like a month of ideas instead of one idea in twenty outfits.

When to use — and when not to

• Use it when long-form already existsa blog post, case study, webinar recap or newsletter with genuine substance; repurposing multiplies what you invested there.
• Use it for the company page and your own professional presenceLinkedIn, X/Twitter, Instagram captions; the recipe fits any mix, one platform or four.
• Use it to escape the content treadmillone strong article per month plus this pipeline beats four thin articles written to feed the feed.
• Skip it for thin source materialan article with one idea yields three posts, not thirty; forcing more produces filler that costs audience trust.
• Skip it for reactive and sensitive postsbreaking-news commentary, crisis response and anything legal/regulatory gets a human author and an approval chain, not a batch pipeline.

The recipe

1. 1.

   Mine the article for angles — not posts. First pass extracts raw material:

   Copy

   Read this article and extract raw material for social posts.
   Do NOT write any posts yet.
   
   [PASTE ARTICLE OR LINK/UPLOAD]
   
   List:
   1. Every distinct claim or insight (one line each)
   2. Every number or statistic, with its context
   3. The most contrarian or surprising statement
   4. 3 questions the article answers that our audience actually
      asks
   5. 2 places where the article disagrees with common practice
   6. Any step-by-step process that could stand alone
   
   Audience: [WHO FOLLOWS YOU, e.g. "HR managers at mid-size
   Indian companies"]. Mark the 8-10 items most likely to stop
   THEIR scroll.
   

2. 2.

   Draft posts per angle, per platform. Second pass turns marked angles into a calendar's worth of drafts:

   Copy

   Using the marked angles, draft a month of posts: [e.g. "12
   LinkedIn posts + 8 X posts"].
   
   Rules:
   - One angle per post — never summarize the whole article.
   - Rotate formats: bold claim, number + context, question,
     mini-story, how-to steps, myth vs. reality.
   - LinkedIn: hook line first (it must survive the "...see more"
     fold), 80-150 words, line breaks for scanability, ONE idea,
     CTA only on some posts — vary between comment prompts and
     [ARTICLE LINK].
   - X: under 280 characters, no hashtag stuffing, at most one
     hashtag where natural.
   - No emojis unless I add them. No "game-changer", "unlock",
     "delve", "in today's fast-paced world".
   - Label each post: Angle # | Platform | Format.
   

3. 3.

   Check variety before quality. Skim the set as a feed, because that is how followers meet it. Good output looks like:

   Copy

   Post 4 | Angle: "62% of exit interviews cite manager, not pay"
   | LinkedIn | number + context
   
   Everyone budgets for counter-offers.
   Almost nobody budgets for manager training.
   
   In the data from 400 exits we reviewed: 62% named their manager
   as the reason. 11% named salary.
   
   The retention budget is funding the wrong fix.
   
   What's your exit data telling you? →
   

   If three posts open with the same rhythm or every hook is a question, fix it at the batch level now — "rewrite posts 3, 7, 9 with a claim-first hook" — before polishing any single post.

4. 4.

   Run the voice pass with real examples. Third pass, and the one that removes the AI smell: paste two or three of your past posts that performed well and say "rewrite the full set to match the voice, rhythm and vocabulary of these examples — keep the angles and structure." Few-shot voice matching beats any list of adjectives about your brand. Then hand-edit the five posts you will publish first; your edits become next month's voice examples.

5. 5.

   Verify numbers, then schedule with intent. Every statistic in every post gets checked against the article — a wrong number in a social post is public, screenshot-able and permanent (pb-09's ritual, applied per post). Then load the set into Buffer: 3–4 posts per week, strongest hooks on your audience's peak days, formats spaced so no two similar posts run consecutively. Leave calendar gaps for timely posts — a fully automated month reads as exactly that.

Make it yours

• ChatGPT or Claude: run the whole pipeline in one chat so the angle list stays in context; Claude's larger context comfortably holds a long article plus voice examples plus the full draft set for the voice pass.
• Buffer: paste the approved set into the queue and use the calendar view as your variety check — the AI assistant can reslant an individual post per channel, but keep angle-mining and voice in your chat pipeline where your examples live.
• The multiplier variant: same recipe runs on a webinar transcript, a conference talk or a customer case study — anything long-form; pb-28 extends this into the full five-asset repurposing chain.
• Team scale: save the three prompts (angles → drafts → voice) as one library entry (pb-05) with your audience line and voice examples pre-filled — anyone on the team can then turn an approved article into an on-voice month, and review shrinks to editing, not authoring.

Quality check

• Every post survives aloneno post depends on the article or another post for its point to land.
• Formats and hooks rotateread the set top to bottom; no two consecutive posts share a shape.
• Numbers verified against sourcestatistics match the article exactly; the platform never forgets a wrong one.
• The voice is yoursa follower would not clock the batch as AI; banned-phrase list respected, your examples audibly present.
• CTAs vary and breathelinks on some posts, conversation prompts on others, nothing on a few; every-post-sells trains people to scroll past you.