 |
|
|
|
|
Inventory Demand Forecasting By Richard Vaccaro, Valogix, Inc. In today's dynamic business climate, companies are looking for simple ways to reduce waste and improve profitability. They go through many business cycles or phases from start-up to maturity to exit. New technology initiatives must be affordable, easy to implement and have a high measurable rate of return. This is also true of how companies approach their inventory planning – from seat-of-the-pants guessing and team collaboration, to spreadsheets that invoke more team collaboration, to forecasting software and possibly even more team collaboration. Many software solutions like accounting packages or ERP systems come with some level of inventory management or inventory control functionality, but rarely do they include forecasting or demand planning. There are stand-alone forecasting packages that offer dozens of statistical models that allow the user to select the one or ones they want to use, and some even provide automated forecasting in which the solution uses a best-fit approach. The business problem However, the question becomes a matter of how to use or interpret the results of forecasting to establish inventory stocking levels and effectively plan. Relying on forecasting systems with a 40%-60% accuracy level at best means exactly what when it comes to purchasing and producing effectively? Also lacking anything else, what generally happens is that forecasts are mistakenly used to set inventory levels or production requirements, min and max levels, and ultimately purchasing and production quantities. This approach is fine if only a few items are involved, but it can be a monumental task for hundreds or thousands of items. The result of this approach in general is either stock-outs and/or excess inventory both of which are costly to the company. Interpreting forecasting results When you have a product with consistent demand and years of history, forecasting tends to be relatively straightforward. However, what about products with extremely random demand, or new products with limited history, or items with infrequent demand? How do you forecast for them? This problem is particularly pronounced with high tech products that may have a sales life as short as a few months. Here are two typical graphs of demand history. The one on the left has a random demand pattern and the one on the right shows a relatively constant demand stream.  In both cases, the forecast trend line may suggest the forecast ramping up slightly at perhaps 110 in the first forecast month, 115 in the second and 120 in the third. Depending on the forecast model chosen, a random demand pattern may cause you to carry more inventory when you don't need it and not enough at times to meet a spike in demand.  Other factors in addition to a forecast should be considered when deciding how much to stock. One factor is Service Level. This can dramatically affect your investment, depending on the service levels you want to provide, as illustrated in the graphs above. Other factors include budget, carrying costs, planning horizon, lead times and many more, such as:
We recommend that you consider forecasting as an important first step but not the only step in inventory planning. The entire inventory planning process must work smoothly and be in sync from forecasting to setting stocking levels to calculating replenishment levels to purchasing and production. Our next article will describe how to use these other factors in concert with the forecast to improve your inventory and increase customer satisfaction. |
|
|
|
1-866-609-6146  |
