This article was originally published on the K3S Blog and is shared with permission by the author.
Somebody once told me that if you’re feet were in a fire, and your head was in a freezer, on average your body temperature would be about right.
But it wouldn’t begin to tell the story of what was going on or your comfort level. Averages can be misleading.
I was recently planning a trip to Las Vegas. 6 months ahead of time, I looked up the average temperature to get an idea of what I should expect. For February, the average high is 70 degrees. However, the record high is 90 and the record low is 16! When it came time to decide whether to pack shorts or jeans a parka, I needed something, far more accurate than the average temperature.
So as the trip approached, I switched from simply looking at the average temperature to the weather forecast. Why? Because weather forecasts factors in much more than a simple average. They follow trends, current weather patterns, and a host of other factors.
This is just as true when it comes to forecasting your product demand. If you’re replenishing inventory and placing orders simply based off of an average, you’re looking at a very small part of the total story. And the part you’re missing could be costing you millions in sales and carrying costs.
A good product forecast takes into account multiple variables such as:
You may have sold a product for 5 years, but suddenly, demand has started to grow. How will a mathematical average react? A standard average will constantly lag behind the trend. If the trend is for greater demand, you will start losing sales. If the trend is for decreasing demand, you’ll be sitting on extra inventory and exposing yourself to greater spoilage and obsolescence.
Imagine 2 products with an average of 50. One product is fairly stable. Looking back historically, it has had a low demand of 43, and a high of 52. On the other hand, you could have a product with the same average of 50, yet has a low of 20 and a high of 100! These items would need to be handled completely differently. Simply knowing that they have an average of 50 is very little help in identifying the right strategy.
It makes sense that you would need more chicken soup in the winter than in the summer, so why would you place an order in November or December that is based on demand from September or October?
Conversely, when you are ordering that same soup in March or April, we don’t want the average to be based on demand from January. In either case, this average demand offers very little help in predicting what you need now. Your demand forecast needs to account for seasonality so that you don’t order off of an artificially high or low demand average at exactly the wrong time.