The importance of a high forecast accuracy in today’s supply chains is higher than ever before – or if not having a high forecast accuracy, at least having a stable and unbiased forecast or forecast range. At the same time, the opportunities of applying more advanced forecast algorithms to the demand planning process have also increased. But how do we know that we are improving the forecast accuracy with a changed statistical model or with the promotion plan that we are currently running?
In this article, we are explaining how applying forecast value-add measures linked to the forecast accuracy of the different building blocks of the demand planning process can help organisations improve their forecast accuracy, increase the control of their forecast and streamline their process.
With the forecast value-add you can measure the value gained of the process steps
In the demand planning process, the different process steps are together constructing the final forecast. The forecast is constructed by reviewing and evaluating the product mix, applying statistical methods, incorporating seasonality knowledge based on historical consumption, collecting input from various stakeholders as sales and finance and adding on extra promotions.
It can be a challenging task to evaluate the input from the different areas, especially when we may miss forecast accuracy measures on the detailed level or if we are only able to compare the final released forecast with the historical actuals. By measuring the incremental shift of the forecast accuracy in each step of the demand planning process, we can evaluate the process steps to assure that the process is increasing the forecast accuracy and runs efficiently.