As of 13 August 2026, AI can build a sales forecast in Excel.

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 your CRM and Excel, export the current opportunity pipeline with opportunity name, owner, stage, value and expected close date, and save the export without changing its source values.
   2. Gather the latest historical sales by month or quarter, the definitions of your sales stages, and any documented conversion rates, seasonality or exclusions.
   3. Paste the prompt into a chatbot and replace each bracketed slot with your business details, forecast period, historical data, pipeline data and rules.
   4. Upload the CRM export and historical sales file if the chatbot supports files, then ask it to create the workbook using only those supplied records.
   5. Open the generated workbook in Excel and compare the pipeline rows, historical totals, stages, dates and values with the source exports.
   6. Open the Checks and Assumptions worksheets, fix every flagged data or formula issue, and record the source for each probability, exclusion and timing assumption.
   7. Ask your sales manager or finance colleague to compare the commit, upside and total forecast with current deal knowledge before sending or using it for planning.

   Prompt

   Build a sales forecast workbook in Excel from the data and rules below.
   
   Business: [business name and what it sells]
   Forecast period: [start date] to [end date]
   Currency: GBP
   Historical sales data:
   [paste dated historical sales by month or quarter]
   Current pipeline data:
   [paste one row per opportunity with opportunity name, owner, customer type, stage, value, expected close date, and any known probability]
   Sales process:
   [list the stages in order and what each stage means]
   Forecast rules:
   [state whether to use historical conversion rates, assigned probabilities, weighted pipeline, a commit and upside view, or another method]
   Known constraints:
   [list seasonality, capacity, contract timing, exclusions and any deals that must not be counted]
   
   Create an .xlsx workbook if your environment supports file creation. Otherwise provide the exact worksheet layouts, formulas and charts needed to recreate it in Excel. Use only the supplied data. Do not invent opportunities, values, dates, conversion rates or assumptions. If a required field is missing, mark it as missing and state how that limits the forecast.
   
   Include these worksheets:
   1. Read me, explaining the forecast method, data date, assumptions, exclusions and limitations.
   2. Clean pipeline, with one row per opportunity and separate columns for source values, calculated probability, weighted value, forecast period and data-quality flags.
   3. Historical results, preserving the supplied historical figures and calculating period totals.
   4. Forecast, showing unweighted pipeline, weighted pipeline, commit and upside by month or quarter as appropriate.
   5. Assumptions, listing every editable assumption and its source.
   6. Checks, testing for duplicate opportunities, missing values, dates outside the period, invalid stages, formula errors and reconciliation between detail and summary.
   
   Use transparent Excel formulas rather than hard-coded summary numbers. Add a simple chart comparing historical results with the forecast. Keep the workbook usable by a non-technical sales manager. At the end, state which figures need checking by a sales manager or finance colleague before the forecast is used. This is not professional advice, and it must not be presented as a guarantee of revenue.

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