The figure above represents the reality of forecast errors faced by typical supply chains. Here’s a summary of what the figure is expressing:
- Forecast errors less than or equal to 20 percent – Items that have high levels of consistent demand are pretty easy to forecast with relative accuracy. The SKUs’ coefficients of variability (COV) will be low, so almost all the forecasts represent relevant information. Unfortunately, in many organizations, the percent of SKUs in this category is very small—not infrequently as few as six percent.
- Forecast errors between 20 and 50 percent – The next group of SKUs will have forecast errors ranging between 20 percent and 50 percent. These probably are items that move fairly steadily and also probably have either shorter average lead times or more frequent replenishment than other the remaining SKUs in the portfolio.
- Forecast errors greater than 50 percent – Typically up to 50 percent of SKUs are lower velocity items which, in turn, also tend to have higher—much higher—coefficients of variability. This makes forecasting accurately almost impossible, regardless of the strength of the underlying forecasting algorithms.
[The numbers of SKUs in the figure is based on an example total of 10,000 SKUs.]
Too much and too little
Of course, this little diagram helps us also see readily why the typical supply chain ends up with bimodal inventory. Bimodal inventories are characterized by the all too typical complaint: “We’ve got too little of the right stuff, and too much of the wrong stuff.”
Relevant information versus irrelevant information
It’s time to face the reality!
Look at the figure above one more time, and consider this:
- For the six percent of SKUs, those with 20 percent or less in forecast error rates, we can easily see that 80 percent or more of the forecasts match actual demand. That means that 80 percent or more of the forecast is relevant. A relatively small portion--only 20 percent or less of the forecast--represents irrelevant information.
- For the middle 44 percent of SKUs, those with 20 to 50 percent forecast error rates, we can readily understand that the forecasts are constituted of 50 to 80 percent relevant information.
- For the remaining 50 percent of SKUs, 50 percent or more of the forecasts are constituted of irrelevant information—the error portion of the forecast.
Simply put, the portion of the forecast demand that matches the actual demand is relevant information. However, that also means that the portion of the forecast demand that does not match actual demand is, in reality, irrelevant information.
As supply chain managers and executives, we see our supply chains relying on large volumes of irrelevant information day after day, month after month. We also see that our supply chain performance, generally speaking, is not improving day-to-day, or month-to-month.
Why do we do this?
Confessing the reality
My observation is that we frequently persist in such folly simply because we won’t change our language. We call data “data,” and we never distinguish between what is relevant in our information flow, and what is irrelevant.
We also fail to connect the fact that, in our supply chains, when we act upon irrelevant information, we end up doing the following:
- Producing irrelevant products—products not required our customers—thus consuming capacities that might otherwise be used to produce relevant products
- Shipping irrelevant products—thus consuming our logistics resources that might otherwise be used for the transportation of relevant materials
- Buying irrelevant products—thus consuming our working capital that might better be invested in flow of relevant materials
- Storing irrelevant products—thus consuming our storage resources that would be much more profitable if they were storing relevant materials
When we start seeing and confessing the reality of our situation, chances are we will start looking in a fresh, new direction for answers.
Just think about this:
How much of your supply chain is clogged with the flow of irrelevant information and irrelevant materials?
If you’d like to talk about this, we would like to hear from you. Please feel free to leave your comments below, or contact us directly, if you would prefer.