As of 13 August 2026, AI can summarise your monthly business performance.

What to actually do

1. Use a tool built for this

   The route this page recommends

2. Do it yourself

   Second choice

   A chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.

   How to actually do it

   1. Open the source systems and export the month’s relevant sales, costs, orders, customers, cash or operational KPI data into a spreadsheet, keeping the original column names and date fields.
   2. Add a separate sheet containing the previous comparison period, budget or target figures, KPI definitions, currency and any known one-off events.
   3. Remove passwords and unnecessary personal data, then check that the dates, currency, units, duplicate rows, missing values and source totals are clear.
   4. Open a chatbot or Julius AI, attach the prepared spreadsheet, paste the prompt, and replace the bracketed instructions with your business name, period, comparison basis and KPI list.
   5. Check the generated totals and percentage changes by calculating them from the spreadsheet, and compare each KPI with the source system or approved report.
   6. Add the business context the data cannot show, correct any unsupported explanation, and mark hypotheses as hypotheses before sharing the final summary with your colleague or decision-maker.

   Prompt

   Summarise the attached monthly business performance data for [business or team] covering [month and year]. Use only the supplied data and do not invent figures, explanations or missing context. First check the column names, date range, currency, missing values, duplicate rows and whether totals reconcile. State any data-quality problems before giving conclusions.
   
   Compare the current month with [previous month, same month last year, budget or target, if supplied]. Report the supplied KPIs: [list KPIs]. For each KPI, show the current figure, comparison figure, absolute change and percentage change only when the comparison is valid. Keep units and currency clear.
   
   Produce:
   1. A short executive summary of the main movements.
   2. A table of KPI results and comparisons.
   3. The strongest positive and negative changes, with the underlying figures.
   4. Patterns or possible explanations supported by the data, clearly labelled as hypotheses rather than facts.
   5. Risks, anomalies and questions that need a human answer.
   6. Three practical follow-up actions based only on the evidence.
   
   Do not treat correlation as causation. Do not give financial, tax or legal advice. If a requested calculation cannot be made from the data, say what is missing. End with a verification checklist listing the source totals and calculations I should check before sharing the summary.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

3. Hand it to a person

   The distant third

   A person who owns the outcome does this end to end, worth it when the failure is dear.