As of 13 August 2026, AI can only partly calculate your customer lifetime value.

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 30 minutes until you can act on the result.

   How to actually do it

   1. Open your customer, order or subscription system and export a dated table containing customer or cohort identifiers, revenue, refunds, costs, cancellations, retention and acquisition cost where available.
   2. Open the export in a spreadsheet and remove test accounts, duplicated transactions and clearly identified internal or free customers, recording each exclusion in a separate notes sheet.
   3. Add the business context to the prompt, including whether the calculation should use revenue, gross profit or contribution after variable costs and acquisition cost.
   4. Paste the cleaned table into the prompt and ask the AI to calculate historical, margin-based and cohort versions only where the data supports them.
   5. Compare every input, total and segment count in the response with the source export, then recalculate at least one displayed formula in your spreadsheet.
   6. Check the retention, churn, gross margin and forecast-horizon assumptions against your current finance and customer reports, and ask the AI to rerun the result when any assumption is wrong.
   7. Send the final table of definitions, assumptions, results and sensitivity cases to the finance or commercial owner before using it to set acquisition budgets, prices or targets.

   Prompt

   Calculate customer lifetime value from the data below. Use pounds sterling and show every formula, intermediate figure and assumption.
   
   Business context:
   - Business model: [subscription, repeat purchase, contract or other]
   - Customer definition: [what counts as an active or retained customer]
   - Analysis period: [start date to end date]
   - Preferred forecast horizon: [state a period, or ask me to compare reasonable horizons]
   - Acquisition cost treatment: [include or exclude customer acquisition cost, and explain why]
   
   Data:
   [Paste a CSV or table containing customer or cohort data. Include dates, revenue, refunds, variable costs, gross margin or gross margin percentage, orders or subscriptions, cancellations, and acquisition cost where available.]
   
   Instructions:
   1. Do not invent missing figures. List missing fields and explain how each affects the result.
   2. Clean and describe the data before calculating anything, including duplicate records, refunds, churn and incomplete customer periods.
   3. Calculate a simple historical customer lifetime value and a margin-based version where the data supports it. If a cohort or retention model is appropriate, calculate that separately.
   4. State the exact definition of lifetime value used, including whether it is revenue, gross profit or contribution after acquisition cost.
   5. Separate observed figures from estimates and assumptions. Do not treat an incomplete customer history as a completed lifetime.
   6. Show results by relevant customer segment or cohort if the data supports this, and explain why segments differ.
   7. Give sensitivity cases for the key assumptions rather than selecting an optimistic value.
   8. End with a short list of checks I must perform against the source systems before using the result in a business decision. Do not present the result as a forecast with more certainty than the data supports.

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