Forecasting Accuracy in dollars instead of percentages

00:08 Introduction
00:45 What does it mean that a forecast is more accurate?
01:26 Does a forecasting accuracy depend on the forecasting metric that you use?
02:05 We have a vast array of different metrics to use. Which one is the best one?
03:14 Can focusing on the problem of maximizing accuracy in percentages be misleading for supply chain performance?
04:40 How do you measure supply chain performance?
06:14 Does forecasting accuracy in percentages coincide with forecasting accuracy in dollars?
06:47 Example: Supermarket
11:20 So in this example, a 100% accuracy in a forecast results in 0$ in revenue?
12:31 Are we not only looking at the problem in the wrong way, but the wrong problem to begin with?
14:38 Should there be a division dedicated to improving forecasting accuracy in the first place?
17:58 The team in charge of forecast
18:40 So the supply chain performance does not improve with improved forecasts?
21:21 Joannes, how did you come to realize that more accurate forecasts did not result in better performing supply chains for our clients?
25:44 How has Lokad’s product evolved?
29:43 Joannes, what would be your advice for big companies?


The supply chain industry has for decades been in a constant chase of better accuracy of forecasts in a desperate attempt to predict the future. In fact, forecasting is so common in supply chain practices that every large company has its own division dedicated to improving their forecasting accuracy.

Nonetheless, despite the substantial resources and efforts invested in improving accuracy, it has not resulted in better performing supply chains. How can chasing better accuracy be counterproductive? Are we blindfolding ourselves by measuring accuracy in percentages, rather than dollars? We explore this key topic in this episode.

A more accurate forecast simply means that according to a certain metric, there is one forecasting model that will be more accurate than another. It is thus the chosen metric that is the defining factor for a forecast accuracy. However, discussions about what forecasting method to use take away the focus from what decisions must be taken from these forecasts. As a forecast is just an educated opinion about the future, it is the decisions - and decisions alone - that affect the supply chain performance.