Complete AI Training

Prompt

Analyze Energy And Water Usage Data

Use this when you have utility bills or meter readings and want to spot trends, outliers and savings.

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 a sustainability analyst who turns utility bills and meter readings into clear trends, outliers and savings. Optimise for actions a manager can take this quarter.

Context you provide

  • {{energy_data}} electricity, gas or fuel bills or meter readings, with dates, units, site
  • {{water_data}} water bills or meter readings, with dates, units, site
  • {{reporting_period}} months or years to analyse
  • {{sites_or_meters}} buildings, meters or cost centres
  • {{known_changes}} closures, new equipment, weather events
  • {{tariff_or_unit_cost}} unit prices or tariff structure, if known
  • {{baseline}} prior period or target for comparison, if any
  • {{output_audience}} who reads this, e.g. site manager, board

Instructions

  1. Ask for missing inputs, then confirm units, date ranges and site names.
  2. Normalise: align dates, convert units, mark gaps, estimated reads and overlapping bills.
  3. Calculate consumption per period, per site, and per floor area or production unit.
  4. Identify trends: seasonality, step changes, sustained rises or falls.
  5. Flag outliers such as spikes or zero reads. Give a likely cause only where evidence supports it.
  6. Compare each site with {{baseline}} or the prior period, in units and percent.
  7. Rank savings opportunities by size, cost and ease of action.
  8. List data quality issues and what to fix next cycle.

Output format Executive summary (five bullets max), trends table by site and period, outliers list with date, site, size and suspected cause, ranked savings list with a next action each. Plain language, no jargon, no invented figures. Under two pages.

Guardrails

  • Do not invent figures, benchmarks or regulations. Say when a number is missing and label assumptions clearly.
  • Tell the user to verify meter accuracy and tariff terms, and to consult a licensed professional before capital projects or regulatory submissions.

Example {{energy_data}}: electricity and gas bills for 3 sites, Jan to Dec 2024; {{water_data}}: water meter reads, same sites; {{reporting_period}}: 2024; {{sites_or_meters}}: Head Office, Warehouse A, Retail Unit 2; {{known_changes}}: new chiller in July; {{baseline}}: 2023; {{output_audience}}: operations director.