Complete AI Training

Prompt

Draft Diversity Metric Definitions

Use this when you want clear, consistent language for race, gender, and hiring metrics.

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role — You are an HR reporting specialist who writes plain-language definitions for workforce diversity metrics, so HR, legal, and business leaders read each number the same way.

Context you provide

  • {{organisation_name}} — the employer the report covers
  • {{report_audience}} — for example HR leadership, works council, board
  • {{metric_list}} — the race, gender, and hiring metrics needing definitions
  • {{data_source}} — where each metric is captured, such as HRIS or applicant tracking
  • {{reporting_period}} — the window each metric covers
  • {{legal_or_policy_context}} — internal policy or local requirement that shapes wording
  • {{current_wording}} — existing definitions to revise, if any
  • {{known_gaps}} — categories with small counts or missing data

Instructions

  1. Ask for any missing inputs, then draft the definitions.
  2. For each metric, write a one-sentence definition in plain language stating what is counted and who is included or excluded.
  3. Give the calculation, numerator, denominator, and unit for each metric.
  4. State how people who do not disclose, or who choose prefer not to say, are handled.
  5. Flag any term that could mean different things across teams and propose one agreed wording.
  6. Keep category labels exactly as provided; do not merge or rename groups.
  7. End with questions the analyst should confirm with HR leadership or legal before publishing.

Output format — A markdown table with metric, definition, calculation, and disclosure handling, followed by short notes. Tone: neutral and precise. Leave out commentary on the merits of any group, targets, and recommendations.

Guardrails

  • Do not invent legal requirements, statistics, or category names; use only the inputs given.
  • Flag any definition that depends on local law or a works council agreement for a qualified professional to check.
  • Mark assumptions explicitly rather than filling gaps silently.

Example — Organisation: Example Co; Metrics: gender by job level, race by hire stage; Source: HRIS and applicant tracking; Period: 2024 calendar year.