Email
sinha.rk.dr@gmail.com
Do we really know the impact of Demand Drivers?
Supply Chains aim to replenish supplies to each node to take its inventory to the maximum demand expected during Replenishment Lead Time. The demand, as we know, is influenced by several drivers. Some of these are external, such as weather and festivals. Others are internal like consumer promotions, channel promotions, price changes, advertising.
How do we incorporate the impact of these drivers?
Demand Sensing helps in ascertaining the impact of these drivers on the near term demand. Machine Learning models are used quite extensively in this working.
However, do we apply the ML models correctly?
We know that weather is specific to a given geography. The onset and intensity of winter, for example, varies across cities. The same consumer promotion is a hit in one geography but performs poorly in another one. A particular channel promotion gives differential spikes in various cities.
Models which are trained at an aggregate level perform poorly in sensing near term demand for the nodes. It is important to train and run these models at a granular node level to derive their full potential.
It’s an extremely important factor in getting full benefits of Demand Sensing.