📌 Quick Context:
One of the fastest ways to boost AI output quality? Give it a role. When you ask AI to act as a specific expert, it draws from relevant patterns, language, and reasoning styles to deliver more targeted, useful responses.
How to Do It in 3 Steps
Step 1: Assign a Clear Role
Start your prompt with “Act as a…” followed by a well-defined expert or persona. This tells the model what lens to apply.
🧠 Examples:
“Act as a CFO and review this budget for inefficiencies.”
“Act as a UX researcher and suggest 3 usability improvements for this app.”
“Act as a personal trainer and create a 4-week beginner strength program.”
✅ Pro Tip: The more specific the role, the better. Don’t just say “expert” — name the domain.
Step 2: Give Context + a Clear Task
After the role, set the scene. What is the AI looking at, and what’s the goal?
🧩 Example prompts with context:
“Act as a recruiter. Review this résumé and provide feedback to make it more appealing for a senior product manager role.”
“Act as a startup investor. Evaluate this pitch deck and highlight red flags.”
“Act as an AP English teacher. Grade this essay and explain what could be improved.”
⚠️ Common Mistake: Being too vague. Without task clarity, responses get generic.
Step 3: Iterate with Role Continuity
Keep the same role for follow-ups. This maintains coherence and lets you go deeper.
🔁 Advanced example:
Prompt: “Act as a climate policy advisor. Draft a one-page brief supporting carbon pricing.”
Follow-up: “Now, act as the same advisor, but rewrite it for a skeptical business audience.”
Further: “Compare this brief with current EU carbon policy, and point out alignment or gaps.”
💡 Insight: This is especially powerful for simulations (e.g., “Act as a hiring manager interviewing me for X role”).
📝 Recommended Next Step:
Try this format with 3 different roles today. Choose one technical (e.g., “data scientist”), one business-oriented (e.g., “CMO”), and one unconventional (e.g., “stand-up comedian”). Observe how tone, structure, and advice change.
📌 Example Prompt Bank (Swipe These!):
“Act as a DevOps engineer. Review this CI/CD pipeline config.”
“Act as a therapist. Help reframe this anxious thought pattern.”
“Act as a restaurant critic. Write a review of this menu.”
“Act as a debate coach. Strengthen my argument for AI regulation.”
“Act as a fiction editor. Punch up this dialogue.”
Role-based prompting is like assigning a job to your AI coworker — and the better the job description, the better the work you’ll get back.
