By now you must have seen how Watson* can tell you all about how much better you can do based on past data. Although this might be a useful information in many cases however it is far from certainty. It can only show a trend or a pattern that may or may not happen in future. As the old adage goes: you cannot drive a car by looking at your rear view window. Typical information from analytic engines shows you the likelihood of a trend. For example, the customers in NE of USA buy 10% more grey coats in winter than NW of the country. In contrast to this approach of relying on past patterns, Supply Side analytics relies on modeling of the future and based on these models it can be a precise predictor of how to run the supply chain. Models, like any other mathematical equation, can be used to plug in the numbers and get results that are predictable and accurate. When a space craft is launched to moon or Mars, its entire journey is defined by an accurate model that predicts how it gets there and where it lands. To this end, we can model supply chains to see exactly how they will behave and where the weak points and risk areas, that might cause issues in future, are.

 

Even more interesting, in supply side Predictive Analytics, is the fact that we can observe potential patterns that seem to be forming in the future. For example, every time the plan runs for a solution, it might highlight certain weaknesses in the supply chain such as low inventory points, long lead-times, single source materials etc. These future pattern are then used to correct the situation before they cause issues. As you can see the future patterns are a much better predictor than past patterns and more reliable source of information. To learn more on this topic click on this link.