The other day I was thinking about the weather forecast, and its inherent uncertainty and how this relates to supply chain forecasting uncertainty. I was reminded of what someone once told me was the true meaning of “a 30 percent chance of rain.” It means this:
"There’s a 30 percent chance 100 percent of us will get wet; or there’s a 100 percent chance that 30 percent of us will get wet; or there’s a 70 percent chance the weatherman’s 100 percent wrong.”
Of course, this humorous depiction of weather forecasting is nearly 100 percent accurate. Here’s how the weatherman comes up with that “30 percent chance of rain” in his forecast.
The meteorologists divide the forecasting area into 100 equal geographic portions—a grid.
However, since weather has so many variables—many of them unknown (more on that later)—the meteorologists cannot predict precisely which areas of the grid might receive “measurable precipitation.” Given the expected size of the weather system and its general direction, they can “guess-timate” about how much of their forecasting area might be affected by the approaching weather system.
The forecasters may, therefore, be able to predict with considerable certainty that “n percent” of the forecast area is likely to get rained upon, but they cannot predict precisely which “n percent” will be the recipient of the rain.
As a result, it may be truly said: a 30 percent chance of rain means that there’s a 30 percent chance that 100 percent of the area will get rain; or a (near) 100 percent likelihood that 30 percent of the area will get rain. There is also, however, a 70 percent chance that none of the forecast area will get rained upon.
The Need is Not for More Powerful Computers
Back in the mid-to-late twentieth century, meteorologists made the bold claim: “Give us powerful enough computers and we will be able to forecast the weather with 100 percent accuracy.”
Since that time, computers and computing science have worked together to prove one thing to meteorologists: the problem is not having enough computing power; the problem is that we don’t understand all of the factors at work in creating and changing weather patterns.
The Butterfly Effect
“The term butterfly effect comes from the suggestion that the flapping of a butterfly's wings in South America could affect the weather in Texas, meaning that the tiniest influence on one part of a system can have a huge effect on another part. Taken more broadly, the butterfly effect is a way of describing how, unless all factors can be accounted for, large systems like the weather remain impossible to predict with total accuracy because there are too many unknown variables to track.”
Consider Your Supply Chain
All that we have said so far regarding weather forecasting has precise parallels in any forecasting methods you may seek to apply to your supply chain. Instead of a geographic area divided up into 100 parts, you might have a hundred (or more) outlets for your products, or you might have hundreds of different makes, models, styles or colors to manage in your supply chain.
On the aggregate level, it may be relatively easy and (more or less) accurate to predict, for instance, “we are likely to sell 30,000 units from our XL1000 product line over then next 90 days.”
However, predicting which models, styles or colors among those 30,000 units will sell in which of the 100 outlets is much, much more difficult to predict. And, because of “the butterfly effect,” no one in your supply chain can even know all the factors involved in creating the specific demand that leads to the movement of a specific model, style and color off the shelf at a particular outlet. And, if someone did happen to know that “a butterfly flapping its wings” in South America was going to affect demand in Akron, Ohio, they would not know the precise value to attach to that change in the demand forecasting formula.
Meteorologist Change Their Focus
You may have noticed, as I have, that over the last couple of decades the weather folks on TV and radio—and those at the National Weather Service, as well—have largely surrendered on the idea of ever being able to predict the weather with 100 percent accuracy.
Instead, they have wisely sought to narrow their focus to those things that are most important. For them, that appears to be saving lives and reducing damage to property by being able to report up-to-the-minute details on weather that is likely to create such losses. They have created new mechanisms and new alliances (with “weather watchers,” “storm chasers,” and broadcast and other media outlets) to capture the effects and warn people who are likely to be affected by “breaking weather.” (For supply chain managers, read: “actual demand.”)
Wisely, meteorologists have chosen agility and responsiveness to changing demand in near real-time over the unfruitful pursuit of “perfecting the forecasting mechanisms.”
Beware the Butterflies
Our supply chains will always be subject to butterflies.
And, there’s a reason the weatherman (I use the term generically, and mean no offense to all those fine women in the position of weather forecasting and broadcasting) predicts only a 10, 20, 30 (and so forth) percent chance of rain. I’ve never heard a weather report that said, for example, “And, there’s a 32.8 percent chance of rainfall today….” Nevertheless, the buyer needs a specific number for ordering product, and the manufacturer must make a discrete number in its production run. The supply chain doesn’t operate on “about n units is what we think we’ll be needing.”
Because butterflies will always be affecting our supply chain—because we can never know all the factors that affect demand or how to quantify precisely the factors we do know—it is wiser to focus our efforts on supply chain agility, reducing the length of our replenishment cycles, making our production and transfer batch sizes smaller, and providing more end-to-end visibility of actual demand all across the supply chain.
The most successful and profitable supply chains buffer against demand variability with capacity, not inventory, and work to balance the flow across the supply chain, not balance stocks on-hand.
We would like to hear what you have to say on this matter. Please feel free to leave your comments here, or to contact us directly.
