Serving up greater responsiveness and less bias

The project

With the desire to improve the foundation for their demand planning process, this global food manufacturer set about creating a stronger, smoother solution that could handle both baseline and promotion demand as well as statistical forecasting.

The new process was built on simple forecasting concepts. A segmentation model helped define how different products should be forecasted and on which product planners should focus their time.

A number of demand planning key figures, both for baseline and promotion demand, were incorporated into the demand planning solution, collectively resulting in a markedly improved consensus forecast.

The number of manually forecasted products was significantly reduced – instead these products were now handled through statistical forecasting.

The impact

• 50% reduction of forecast bias.
• 22% fewer forecast errors.
• Reduction of manually forecasted products from 60% to 25%, creating huge manual task time savings.
• Significant reduction in manual data loads and data preparation in Excel, equivalent to 3 work days/month in the supply chain planning team.
• Improved data and process transparency through one set of numbers: shared data model between sales, demand planning and S&OP.