One Prompt, Three Audiences
You wrote a good explanation. Clear, well-structured, says what needs saying. Then someone asks you to send it to a different audience — the exec team, the client, the new hire — and you open a blank document and start over.
You just rewrote 80% of the same information because 20% of it needed a different wrapper.
This happens because you are treating adaptation as a creative task when it is a mechanical one. The information does not change. The frame changes: what you lead with, what you skip, what you explain, and how much patience you assume. Those are four variables, and once you name them, the model handles the rest.
Four prompts.
1. Define the reader, not the tone
The reader for this version is [role/title]. They care about [what], they already know [what], and they will stop reading if [condition].
Do not change my tone. Change what you lead with, what you explain, and what you skip. The information stays the same; the frame changes.
"Write it for executives" produces a parody of executive communication — shorter, vaguer, and full of words like "strategic" and "leverage". That is not adaptation. That is costume.
When you define what the reader knows, what they care about, and what makes them stop reading, the model makes structural decisions — lead with the outcome, skip the methodology, explain the risk — instead of cosmetic ones.
2. Make it show its choices
Before you write, tell me: - What are you leading with, and why? - What are you skipping entirely, and why? - What are you explaining that the original does not, and why?
If any answer is "because it is for executives", try again with a concrete reason.
This is the prompt that catches lazy adaptation. Without it, the model makes changes that feel right but cannot be defended — and you cannot tell whether the adaptation is good or just different.
The ban on "because it is for executives" is the lever. That phrase is a shortcut to generics: shorter sentences, bigger words, less detail. Forcing a concrete reason — "leading with revenue impact because this reader approves budget" — produces adaptation that actually fits.
3. Preserve what made the original work
Read the original again. Name the three strongest sentences — the ones that would lose the most if removed.
These must appear in every version, adapted but not diluted. If a sentence cannot survive the adaptation, tell me what would be lost and suggest how to preserve the point in a way this reader can receive it.
Adaptation without a preservation rule produces a worse version of the original every time. The model dutifully adjusts everything, including the parts that were already working, and the result reads like a competent summary with no edge.
This prompt anchors the adaptation. The three strongest sentences are the skeleton — everything else can move around them, but they stay.
4. Put both versions side by side
Show me the original and the adapted version in two columns. For each paragraph, annotate what changed and why in one sentence.
If any change cannot be explained in one sentence, it was not an adaptation — it was a rewrite. Undo it and try a smaller change.
The side-by-side is the quality control. You see instantly where the model rewrote instead of adapted — and those spots are almost always where the original was fine and the model changed it anyway, because changing things is what it was asked to do.
The one-sentence rule is strict for a reason. Good adaptation is small: move the conclusion up, cut the methodology, add one sentence of context. If the change needs a paragraph to explain, it is not adapting — it is editorialising.
Why adaptation beats rewriting
Rewriting wastes the hardest work: figuring out what to say. The thinking, the structure, the examples — all of that stays. What changes is the presentation layer, and the presentation layer is the part the model is actually good at.
One piece of thinking. Three versions. Zero blank pages. The information stays the same because it was right the first time. The frame changes because the reader changed, and the frame is all the reader sees.
The four prompts are in the carousel on Instagram.
Want the calm version of AI news like this, once a week? Subscribe to the Sharp AI Hub newsletter →