There is a strong relationship between erratic demand and forecast accuracy. Demand volatility does not necessarily mean the demand is unpredictable.
For example, the calculated demand volatility fro a product could be the result of a highly seasonal demand profile. Such highly seasonal demand can look nice on a graph and could be easily predictable.
In summary, volatility is not necessarily a bad thing. Growth causes volatility, so do seasonal and cyclical demand swings. There can also be other reasons - promotions, item life cycle stages (phase in and phase out) price point reactions, and changes in logistic schedules.
Coporate management can try to stay away from some suboptimal business practices and as a result cause a reduction in demand volatility. They can promote product to customers to increase demand for a period time that can artificially smooth the normal seasonal pattern. This is only one reason why it is important to exclude promotional demand from the seasonal qualification and consider only normal or replenish demand.
Our lives are essentially seasonal and that is something we have to live with. Thanksgiving, Christmas and other holidays do not occur each month. Customers like to go to the beach on sunny summer days and more of these days occur during certain times of the year.
Forecasting accuracy is an expression of how we can predict actual demand - erratic or not. You may be able to create a very accurate forecast from a volatile demand. this can be done be performed through good modeling and diligent forecasting and maintenance.
Balancing service level goals across multiple channels of demand can be a daunting proposition. But with Replenishment, we can easily acheive this. Managing our distribution center efforts has really turned from an art to a science with Replenishment.