Audience rewrite prompt builder
The same facts land very differently with a teenager, a busy executive, and a domain expert. A good rewrite prompt does not just say “make it simpler” — it specifies the reading level, the background knowledge the reader already has, a clear jargon policy, and any cultural context that shapes word choice and examples. This builder assembles those instructions into one reliable prompt.
Why “make it simpler” is not enough
If you send a model a document and ask it to “simplify this for non-technical readers”, you will get a different kind of simplification every time. The model has to guess: How simple? Which terms need defining? Should it use analogies? Should it keep the existing examples or replace them? For a single document this is inconvenient; across a hundred-article content library it produces inconsistency that erodes credibility.
The prompt this builder produces solves that by specifying all the constraints the model would otherwise guess at — once, in one place, reusably.
How it works
You describe the audience and pick a reading level, then declare how to handle jargon and add optional cultural notes. The tool orders these into a structured prompt with explicit rules: preserve meaning, add no new claims, keep roughly the same length, and return only the rewritten text. Because every constraint is stated, the model produces consistent output you can apply to an entire document set.
Reading levels as a calibration tool
The reading level selector is the most direct calibration mechanism in the prompt. A rough guide to what each level implies:
| Level | What it means in practice |
|---|---|
| Grade 5 | Short sentences, common words, no assumed knowledge, concrete examples |
| Grade 8 | Standard newspaper level; some technical terms if immediately defined |
| Grade 12 | Comfortable with complex sentences and light domain vocabulary |
| Undergraduate | Can follow structured argument; domain basics assumed |
| Expert | Full domain vocabulary; no need to define standard terms |
The level you choose determines the density of vocabulary, sentence complexity, and how much assumed context the model uses. For health, financial, or legal content aimed at the general public, grade 8 or below is usually the right target.
Jargon policy options
Remove all jargon is appropriate when the audience has no domain background and technical terms will simply cause them to disengage.
Define jargon on first use is the standard approach for mixed audiences — keeps domain terms for precision but ensures accessibility.
Keep jargon is appropriate for expert audiences where the technical vocabulary is the shared language and removing it would actually reduce clarity.
Pairing jargon policy with reading level is important: removing jargon while targeting an expert level produces awkward circumlocutions; keeping jargon while targeting grade 5 defeats the purpose entirely.
Tips for better rewrites
- Be concrete about the audience. “Busy small-business owners with no technical background” beats “general readers” by a wide margin.
- Match jargon policy to reading level deliberately — they are paired constraints.
- Use cultural notes to lock spelling and idiom. Specify UK or US English and any metaphors to avoid so the output never drifts.
- Add a “do not include” note for any topics or phrasing the original contains that should not survive the rewrite.
- Save the prompt as a reusable profile. Reapply it to every new article to keep one consistent voice across your content library.