Two reasons, and the second one is the one nobody catches. It was cross-posted rather than derived, which means the same text was moved to a platform it was not written for. And every version was generated in a single pass, so all of them inherit the same rhythm and sentence length, which is immediately obvious to anyone who reads more than one.
Both are fixable in the same afternoon.
Cross-posting and deriving are different things
Cross-posting is taking text and putting it in more places.
It reads as displaced everywhere except where it was written, because every platform has a different reading posture. Someone on LinkedIn is scanning between meetings and deciding in one line whether to expand. Someone reading an email opened it on purpose and has already committed. Someone on Instagram is not reading at all until something stops them. The same paragraph cannot serve all three, and a paragraph optimised for one of them looks slightly wrong in the others in a way people feel without naming.
Deriving is taking the argument rather than the text, and rewriting it for the platform. Same point, different opening, different length, different next step. It takes longer than cross-posting and it is the difference between content that travels and content that visibly commuted.
The single-pass problem
This is the subtler one and I did it for a long time without noticing.
Ask for six platform versions in one request and you get six versions of the same shape with the words changed. The same rhythm, the same paragraph length, the same trick in the opening line. Individually each one is fine. Read two of them within a week and the pattern is loud.
The fix is unglamorous: derive them one at a time, each as a separate request, with that platform's rules loaded fresh. It is slower by maybe fifteen minutes and it is the difference between six pieces and one piece six times.
Long-form first, always
The rule that saves the most time is that the long-form piece comes first and everything short is derived from it.
Writing a short post and then trying to expand it produces work with no spine, because the argument was never fully made anywhere. You end up padding. Whereas a complete argument can be cut into almost any shape and each piece still stands, because the thinking was done once, properly, up front.
This is also why the long-form piece is the only expensive step. Everything after it is derivation, and derivation is fast.
The Content Multiplication Playbook
Third of the six playbooks. It produces one finished long-form piece and four to six platform-specific derivatives, each with its own next step attached.
The build.
- Write the long-form piece fully, with the examples in it. This is the only step that takes real time.
- Name the derivatives before you generate anything. Four to six, each with a named platform and a named job.
- Load your context brain plus the platform rules. The platform rules are the part everyone skips: what the first line has to do, how long, what reading posture, what tone is native there and what reads as imported.
- Derive one at a time. Separate requests. This is the step that matters most and takes the least effort.
- Attach one specific next step to each. If you cannot name one, the derivative should not exist.
Step five is the check that stops orphan content, and it fails more often than you would expect. We have written more on the surrounding system in the AI co-building workflow and on building a content schedule without spending all day on it. If the problem is that the writing itself sounds machine-made rather than recycled, that is a different issue and it is covered in how to draft content that does not sound like AI.
The trap
Multiplying something that was not worth writing.
A system that produces derivatives will produce them from whatever you give it, and it does not have an opinion about whether the source had a point. So you can end up publishing six versions of a piece that should not have existed, on a reliable schedule, which is a more efficient version of a problem you had anyway.
The check is the same one that applies to every system in this series. Before the multiplication step, ask whether you would have sent the long-form piece to one specific person you respect. If not, do not fan it out.
Frequently asked
Why does my repurposed content sound recycled? Usually because it was cross-posted rather than derived, or because every version was generated in one pass and inherited the same rhythm and structure.
Should I write long-form or short-form first? Long-form first, always. Short-form derived from a complete argument has a spine. Short-form expanded into long-form does not, and the padding shows.
How do I repurpose content with AI without it sounding the same? Derive each platform version as a separate request with that platform's rules loaded, rather than generating all versions in one pass.
How many pieces should I derive from one article? Four to six is the practical range. Past that you are producing pieces with no distinct job, which is how orphan content happens.
What is the difference between cross-posting and repurposing? Cross-posting moves the same text to more places. Repurposing takes the argument and rewrites it for each platform, with a different opening and a different next step.