0 Replies Latest reply on Jul 6, 2010 12:31 AM by Duncan Klett

    See "In the eye of the supply chain hurricane"

    Duncan Klett Elite

      I saw this article the other day entitled “In the Eye of the Supply Chain Hurricane”. 



      The author, David Blanchard, ends his comments with the following two paragraphs.

      “Given the billions spent on supply chain planning over the last 20 years, why is forecast accuracy still a problem?” Klappich asked the audience, somewhat rhetorically. He cited what he calls the “carpenter analogy”: You can hand a state-of-the-art cutting tool to an average homeowner, but that won’t make him a master carpenter. “We have great supply chain planning tools now,” he noted, “but we’re still learning how to use them.”


      Not surprisingly, then, investment priorities for improving supply chain management find “improving planning processes” at the top of the list, garnering 20% from the respondents. Second was “aligning supply chain management with the corporate business strategy” (11%), and third was “improve supply chain visibility” (10%).


      Why is forecast accuracy still a problem? 


      Compare supply chain forecasting with weather forecasting.  With many years of research, global reporting, and the world’s supercomputers, weather forecasting is getting more reliable.  However, even for the next three days, forecasts are still not great (A recent weekend forecast was two days of rain.  Reality was a half day of rain and a fabulous, mostly sunny day).   My point is weather is a physical process.  Weather forecasting is complicated, but fundamentally ties to predicting those physical processes.


      On the other hand, forecasting demand for a particular product is much more complicated.  Certainly, some portion of the demand can be predicted based on trends, seasons, global economics, and the like.  In addition, some product demand is related to other phenomena.   Consider the daily demand for ice cream and hot dogs which depend upon the weather, which we already know is not perfectly predictable.  Unfortunately, most demand is a function of decisions made by individual people, in response to their specific needs.  Even with perfect information, and over aggregated data, the best forecast is subject to a margin of error.


      Perhaps after “billions spent”, it is time to recognize that a perfect demand forecast is not possible.  If so, what can be done?


      Demand forecasts help supply chain managers plan supply to satisfy that demand.  Certainly, a perfect forecast, available months in advance, would make supply chain management much less risky.  It would only leave supply problems to be managed.


      Considers some opportunities:  If your supply lead time was zero – if you could provide any quantity of any product at any time, would you still need a perfect forecast?  If you did not have to worry about capacity, would you still need a perfect forecast?  Is there a workable balance between a “good enough” forecast and a responsive supply chain?  Maybe it is time to invest in making your supply chain more responsive rather than pouring even more money into improving your forecast.