In chapters 14 and 15, we discussed how forecasting models work. Forecasting engines will populate a forecast baseline, but this is not the end of your forecasting process. As illustrated in figure 16.1, different teams can still enrich the forecast using various insights and sources of information. Because they rely on human judgment, these adjustments are called judgmental forecasts.
Figure 16.1 Example of a demand planning process
As we will discuss in this chapter, judgmental forecasts come with risks and pitfalls. Nevertheless, if done correctly, they should add great value to your forecasts.69
First and foremost, tracking forecast value added (chapter 12) is the cornerstone of any demand planning process using judgmental forecasts. You absolutely need to implement it to enforce ownership and accountability for all stakeholders participating in the demand planning process. Tracking added value will help you to monitor your process, but looking at this metric alone won’t provide advice or guidelines on how to edit the forecast or why some colleagues struggle to add value. This is what this chapter is about.
16.1 When to use judgmental forecasts?
Let’s start by asking ourselves a general question: In which case should planners use their own judgments to edit the forecast baseline?
In general, using judgmental forecasts is a good idea if you can leverage information your forecasting model is unaware of. Here are a few good and bad examples: