Skip to content

demand-forecasting

Predict product demand from historical sales data, identify seasonal patterns, and project forward from the current order pipeline. Works from CSV or spreadsheet exports — no BI tools, ERP access, or specialist software required.

  • Area: Operations
  • Category: operations
  • Level: beginner
  • Input: conversation
  • Output: document
  • Setup: none
  • Roles: Operations Manager

Predicts demand from historical sales data, seasonality patterns, and pipeline data.

How you ask

  • “Forecast demand for next quarter from this sales CSV”
  • “Which of our products have seasonal peaks and when are they?”
  • “Based on our current order book, what do we need to make in the next 8 weeks?”
  • “Set inventory targets for each SKU based on last year's data”
  • “Flag any open quotes that we can't fulfill if they all convert this month”

Inputs

  • sales-data — CSV or spreadsheet with at minimum: date column, product/SKU column, quantity sold column
  • forecast-horizon — How far to forecast (e.g. 'next 4 weeks', 'next quarter'). Defaults to one period matching the input data granularity.
  • method — Forecasting method. 'auto' selects based on data volume and variance. Defaults to auto. (simple-moving-average | weighted-moving-average | auto)
  • sales-data — CSV or spreadsheet with date, product/SKU, and quantity columns. Minimum 24 months recommended.
  • products — Comma-separated product names or SKUs to analyze. Omit to analyze all products in the file.
  • pipeline-data — CSV or spreadsheet with open orders, quotes, or enquiries — including expected close date and quantity.
  • conversion-rate — Historical quote-to-order conversion rate as a percentage (e.g. '65%'). If omitted, the skill will ask.
  • lead-time-weeks — Production or procurement lead time in weeks. Used to flag urgent gaps.

Outputs

  • forecast-table (markdown)
  • pipeline-forecast (markdown)
  • prefer a permanent directory)
  • seasonal-index-table (markdown)