Adding Context: Who, What, Why

Of the three context dimensions — who, what, why — one usually changes the output more than the other two combined. Guess before you tap.

Three dimensions cover most of what a prompt is missing. They map to the anatomy of a prompt you've already seen — context is one organ, and it has three parts: **Audience**: Who the response is for — their role, knowledge level, and relationship to you. **Subject**: What the situation actually is — the facts the model can't observe on its own. **Goal**: Why you're asking — what the answer needs to accomplish once you have it. You don't need labeled sections. A sentence per dimension, woven in naturally, does the job. Audience context is also a close cousin of personas, where you assign the AI a role — that's a whole course of its own.