DAIRY • DEMAND FORECASTING
Dairy Demand Forecasting
A demand-forecasting demonstration built on the MEVGAL sales dataset to support weekly production and stock planning across seven dairy products. Using three years of daily sales data (2020–2022), the analysis developed seven-day-ahead forecasts and translated forecast uncertainty into product-level safety-stock options.
The forecasting approach combined recent sales, weekday patterns, yearly seasonality and points of distribution. All seven product forecasts outperformed the same-day-last-week baseline in the 52-week evaluation. The analysis also examined forecast misses and the trade-off between stock-out protection and excess stock for perishable products.
This was a forecasting demonstration, not a live production deployment. The deck's illustrative weekly plan covers 1–7 January 2023; a pilot would require current operational data, including promotions, prices, stock-outs and product shelf-life information.