As of 13 August 2026, AI can only partly create a KPI dashboard for your UK small business.

What to actually do

1. Hand it to a person

   The route this page recommends

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

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, chat-fluent skill, and roughly 1 hour until you can act on the result.

   How to actually do it

   1. Open the spreadsheet or export that contains your sales, costs, customers, marketing or operational data, and make a copy with personal customer details removed where they are not needed.
   2. Write down your business goals, the people who will use the dashboard, the review frequency and any KPI definitions or targets you already use.
   3. Paste the business context, KPI definitions, targets and a representative data table into an AI chat using the supplied prompt, or attach the copied spreadsheet if the tool supports file analysis.
   4. Ask the model to produce the data-quality findings and KPI specification before accepting any dashboard calculations.
   5. Compare every formula, source field, date range and calculated figure against the original spreadsheet, correcting the model where the source data does not support its result.
   6. Build the checked layout in a spreadsheet or dashboard tool, then test it with a second period of real data and compare each displayed KPI with a manual calculation.
   7. Set an owner and update routine for the source data, and send the finished dashboard to its intended users with the definitions and known limitations attached.

   Prompt

   Create a practical KPI dashboard for my UK small business using the data and context below.
   
   Business type: [describe the business]
   Business goals for the next [period]: [list the goals]
   People who will use the dashboard: [roles]
   Review frequency: [weekly or monthly]
   Available data sources: [spreadsheets, accounting system, sales system or other sources]
   KPI definitions I already use: [paste definitions, or write none]
   Targets: [paste targets, or write none]
   Data: [paste a table or describe the attached spreadsheet]
   
   Requirements:
   1. Do not invent data, targets, benchmarks or definitions. Mark anything missing as "not supplied".
   2. Recommend no more than eight KPIs and separate outcome KPIs from activity or leading indicators.
   3. For every KPI, give its name, purpose, exact formula, numerator and denominator where relevant, source column or field, reporting period, target if supplied, and a plain-English interpretation.
   4. Identify duplicated, ambiguous, incomplete or inconsistent fields in the data before calculating anything.
   5. Calculate only what the supplied data supports. Show the calculation steps and flag figures that need confirmation.
   6. Propose a dashboard layout with sections, chart types, filters and an update process suitable for a small UK business.
   7. Include a short list of decisions the dashboard could support and the questions it cannot answer.
   8. Return the result as: data-quality findings, KPI specification table, calculated results, dashboard layout, update checklist and open questions.
   
   Use British English and pounds where the supplied data uses pounds. Keep the dashboard useful rather than decorative.

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