Why it's worth the line
A model can't write for someone it can't picture. Leave the reader blank and it writes for a vague everyone — which is no one. You get the beige, hedge-everything middle: technically on-topic, useless in the room you'll actually send it into.
The audience is upstream of tone, length, vocabulary, and how much you have to explain. Decide it once and the model stops guessing on all four.
How to catch it in your own prompt
- The prompt could describe ten different deliverables and you'd accept any of them.
- You wrote "make it professional" instead of naming who's reading.
- You're about to rewrite the output to be simpler / sharper / less jargon-y — that edit was the missing audience.
What to write instead
- State the reader in a phrase: their role, what they already know, and what they're skeptical of.
- Add the one thing the reader cares about most, so the model leads with it.
- If the audience is mixed, name the primary one and let the rest read along.
Seen side by side
Write a post announcing our new pricing.
Write a post announcing new pricing for existing small-business customers who are price-sensitive and already trust us.
Why it lands — "Existing, price-sensitive, already trust us" tells the model to reassure, not sell — a completely different post than one aimed at new prospects.
Explain how our API rate limiting works.
Explain how our API rate limiting works to a junior frontend dev who's never hit a 429 before — assume no backend background.
Why it lands — Naming the reader's knowledge level kills the two failure modes at once: condescending over-explaining and jargon that loses them.
Stop fixing this by hand.
Paste a prompt — Meerkat catches this and the seven other weak spots before you send it. Free, no signup.