Does this prompt work with Claude or Gemini?

Yes. It relies on no ChatGPT-specific feature: paste it as-is into Claude or Gemini and the behavior is the same.

Why does the prompt ask one question at a time instead of all at once?

Because a role is defined by digging into it. One question at a time forces you to sharpen each expected outcome, rather than skim a list. Eight focused questions beat a twenty-field form filled in a hurry.

What happens for criteria I haven't decided yet?

The prompt leaves them marked "TBC" instead of inventing them. You leave with an honest scorecard that separates what you have decided from what still needs deciding.

← All prompts

Hiring & People · Updated July 17, 2026

ChatGPT prompt for building a hiring scorecard

This is the prompt we run the moment a hiring decision is made. It interviews you one question at a time, then turns your answers into a scorecard. Its rule: it sets no outcome or criterion on your behalf; anything you haven’t settled stays marked “TBC”.

The prompt to copy

We're hiring a {role} for a {N}-person {type} team. Interview me one question at a time (max 8 questions) about outcomes this person must deliver in year one, then produce a scorecard: mission (1 line), 4–6 measurable outcomes, must-have competencies, nice-to-haves, red flags.

Replace the fields in curly braces (the role, the size and type of team), then answer the questions as they come, like a real conversation.

When to use it

The moment you decide to hire, before writing a single line of the job post. It is the first link in Playbook P7: the scorecard then feeds the job post (H2), the screening kit (H4), and the onboarding plan (H6). Thirty minutes of framing here saves weeks of vague interviews.

What you get back

A mission summed up in one line.
Four to six measurable outcomes for year one.
Must-have competencies and nice-to-haves, clearly separated.
The red flags to watch for in interviews.

Real example: from eight questions to the scorecard

Input: Maya, founder of the Brightside agency (8 people), is hiring a project lead. The prompt asks her eight questions, one at a time. Output, the part that matters:

Expected outcome (year 1)	Measure
Deliver client projects by the agreed date	90% delivered on time
Cut approval back-and-forth	Target TBC
Rebuild the relationship with the 3 key accounts	Monthly check-in in place

Notice the second row: Maya hadn’t set a numeric target, so the AI wrote “TBC” instead of inventing a percentage. That is the prompt working, not failing: the scorecard separates what’s decided from what still needs deciding.

The full scorecard, with the eight questions and the P7 playbook that chains it through to the onboarding plan, ships inside the kit.

Frequently asked questions

Does this prompt work with Claude or Gemini?: Yes. It relies on no ChatGPT-specific feature: paste it as-is into Claude or Gemini and the behavior is the same.
Why does the prompt ask one question at a time instead of all at once?: Because a role is defined by digging into it. One question at a time forces you to sharpen each expected outcome, rather than skim a list. Eight focused questions beat a twenty-field form filled in a hurry.
What happens for criteria I haven't decided yet?: The prompt leaves them marked "TBC" instead of inventing them. You leave with an honest scorecard that separates what you have decided from what still needs deciding.