Tip 01
Give it your context
AI doesn't know you work at a 150-person Series B SaaS company in San Francisco. Tell it. Every time. The more specific you are, the more relevant the output.
❌ "Write a job description"
✅ "Write a JD for a Senior HRBP at a 200-person fintech startup in NYC, reporting to the CPO"
Tip 02
Assign it a role
Tell Claude or ChatGPT who to be. "Act as an experienced CHRO" or "You are an employment lawyer in California" dramatically shapes the tone and depth of the response.
✅ "Act as an experienced HR Business Partner who specializes in tech companies and has coached 50+ managers on performance conversations"
Tip 03
Specify the format
Don't let AI decide how to structure the output. Tell it exactly what you want: a table, bullet points, a numbered list, a Notion-ready doc, a 150-word summary.
✅ "Format as a table with columns: Level, Salary Range, Bonus %, Equity %. Then add a 2-sentence summary below."
Tip 04
Constrain the output
AI will write forever if you let it. Give it limits: word counts, number of items, reading level, tone. Constraints force it to prioritize and produce cleaner output.
✅ "Keep the total under 300 words. Use plain language a new hire would understand on day one. Avoid HR jargon."
Tip 05
Treat it like a conversation
Your first response is a starting point, not a final answer. Follow up. Say "make this more direct," "add a more empathetic tone," or "what am I missing?" Iterate fast.
✅ Follow-ups that work:
"Make it sound less corporate"
"Add a section on legal risk"
"Give me 3 alternative versions"
Tip 06
Ask it to flag gaps
One of the most powerful moves: ask AI to tell you what it doesn't know, what it's assuming, or what questions you should be asking. It turns the tool into a thought partner.
✅ "After drafting, list 3 things you assumed that I should verify, and 2 questions a skeptical executive might ask me."