This is an easy question to answer, but many companies answer it incorrectly. Often, the forecast could be more effective. In some instances, the forecast even impedes a company’s efforts to manage their inventory better.
I have already discussed that forecasting is not inventory management and what the difference is (see – Forecasting … Not Buying It). This post is about how to determine whether your forecast is improving your inventory or making it worse. Just a reminder, it is the forecast error (and your lead time to replenish) that determine how much inventory you need to carry. Therefore, the answer to our question is based on the forecast error.
So, how do we determine if the forecast error is too high? We need a baseline to compare to. Fortunately, this exists and is quite easy to determine. There are many basic forecasts available such as, use the last value, use last year’s value, etc. The most basic statistical forecast is using the mean of the previous values. This, also, can provide us a baseline to use to determine if our forecast is helping or hurting.
Again, the most important component of a forecast is the forecast error. Using the mean (or I prefer a moving average based on the cyclicality of the shipments) we can calculate the standard deviation. If the forecast error that you obtain using whichever forecasting methods you choose is greater than the standard deviation of the moving average, your forecasting is probably hurting your inventory.
I like to keep things simple. For most products, I feel that using a basic moving average is the best forecasting method. There are probably very few products where the use of sophisticated forecasting techniques is necessary. In a few industries and for very few products it makes sense, but check to see whether your use of sophisticated models is helping or hurting your inventory management.
For more, see my articles at http://scientificbusinessmanagement.com/blog
To follow me, hop on my Twitter @ScientificBus
Best regards, Shane.