A chatbot with no role answers like an average of everyone. One sentence that says who it is — and who it is talking to — shifts the vocabulary, the priorities and what it leaves out.
Review this ad script.
You are a senior creative director at an FMCG group who approves TV and social scripts for a snack brand sold in the Gulf. Review this script the way you would before sending it to the client.
The prompt
Copy-ready prompt
You are [ROLE — e.g. a senior creative director at an FMCG group] who [WHAT THIS PERSON DOES ALL DAY]. You are talking to [WHO — e.g. a brand manager with no production background]. Task: [THE TASK]. Judge it the way this person would: [2–3 THINGS THEY CARE ABOUT MOST].
Why it works
Anthropic’s guide says setting a role “focuses Claude’s behavior and tone for your use case” and that “even a single sentence makes a difference.” OpenAI’s guide treats identity as the first section of a well-structured instruction, before the instructions, examples and context.
Use it for
Reviews, critiques and anything where the answer should sound like a specific professional. Name the market too — a Gulf retail buyer and a UK brand manager do not care about the same things.
Sources
- Anthropic — Claude prompting best practices — platform.claude.com (accessed 29 Sep 2026)
- OpenAI — Prompt engineering guide — developers.openai.com (accessed 29 Sep 2026)
More tricks and templates in the Prompt Lab. Want this applied to your brand’s brief? Start a project.