How does your Demand Planning tool handle ‘False Positives’ in New Product sales?

Actual demand for a new product at the granular level has high variability and is difficult to predict. On top of it, there are ‘false positive’ signals from time to time, which could easily be construed as demand acceleration.

One such case is when we sometimes see a large quantity picked up on a day in a specific geography. Should we take it as a signal of demand acceleration? It could very well be, if we look at demand only at an aggregate level. However, for all we know, our competitor may have picked it up in bulk for their lab and consumer level testing!

How do we detect and correct for such ‘false positive’ signals of acceleration? Does the tool have outlier detection capabilities? What about self-correcting mechanisms?

The faster we detect and auto-correct such instances, the better our Demand Plans will be.