As of 13 August 2026, AI can only partly create best-case and worst-case sales forecasts.

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, power-user skill, and roughly 1 hour until you can act on the result.

   How to actually do it

   1. Open your CRM or sales spreadsheet and export the current pipeline with opportunity ID, stage, value, expected close period, probability, next step and last activity.
   2. Gather historical sales results for comparable periods, including closed-won and closed-lost opportunities and any recorded conversion or slippage information.
   3. Write down the forecast period, currency, sales constraints and the assumptions your team normally uses for stage conversion, delayed closes and new opportunities.
   4. Paste the business context, historical data, pipeline, constraints and assumptions into the prompt, replacing every bracketed slot and removing any field you cannot support.
   5. Ask the model to produce the two scenarios and flag missing or conflicting inputs instead of filling them with estimates.
   6. Copy the forecast tables into a spreadsheet and recalculate every total, probability treatment and scenario adjustment against the CRM export.
   7. Ask a sales manager or colleague to challenge the assumptions and opportunity treatment, then amend the forecast and record which assumptions changed.
   8. Send the final scenarios with the assumptions, data-quality warnings and confidence limits to the people responsible for targets, staffing or cash planning.

   Prompt

   Create a best-case and worst-case sales forecast from the information below.
   
   Business context: [what we sell, sales cycle, sales period and relevant market context]
   Forecast period: [period]
   Currency: GBP unless stated otherwise
   Historical sales data: [paste the relevant historical results, including the period covered]
   Current pipeline: [paste one row per opportunity with opportunity name or ID, stage, value, expected close period, probability if used, next step, last activity and any known risk]
   Known constraints: [capacity, territory, stock, delivery, pricing or other constraints]
   Assumptions: [conversion rates, slippage, new opportunities, expansion, churn or other assumptions]
   
   Use only the information supplied. Do not invent opportunities, values, probabilities, historical results or market facts. If a required input is missing, identify it and show how it affects the forecast rather than guessing.
   
   Produce:
   1. A concise list of data-quality problems and missing inputs.
   2. The calculation method and every assumption used.
   3. A best-case forecast and a worst-case forecast for the period, with totals in GBP.
   4. A table showing each opportunity, its treatment in each scenario, and the reason.
   5. A comparison of the scenarios, including the opportunities and assumptions that create the largest difference.
   6. A short list of actions that would most improve forecast confidence.
   
   Keep reported facts separate from assumptions. Show the arithmetic clearly enough for a sales manager to reproduce it in a spreadsheet. Do not present either scenario as a prediction or certainty.

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