How Accurate the Forecast Needs to Be?
Why Forecast
First , before summer ends, I wish everyone had a nice summer vacation. If you and your family went to a nice location in this summer that you have never been before, I bet one of the most important steps of your preparation is to check the local weather forecast. Because the forecast will help you to determine what date of travel is ideal, how much clothes to bring, and what service level of hotel and transportation to book etc. …
When you are running a supply chain organization, just same as you are planning a vacation trip, forecast is a dominant factor to make critical decisions such as the time and quantity of purchase or manufacturing, inventory and storage needs, and costs of goods as well as the associated transportations.
How accurate the forecast needs to be?
While our supply chain is relying on the forecast, of course, we expect the forecast to be as accurate as possible, so we can make wise and correct decisions. But I bet that you have heard these similar comments before:
“Nobody has a crystal ball…”
“After all, a forecast is a forecast…”
“Forecast cannot be 100% accurate, it is by definition a prediction…”
So my question is if you know the forecast is wrong any way, how accurate it needs to be and how much efforts should go into improving its accuracy.
My Caribbean Experience
We took a trip to Puerto Rico and Virgin Islands this summer. The beautiful ocean scenery took our breath away. We had a great time in snorkeling, kayaking, rock climbing and hiking in the woods, enjoying a wide variety of fun activities.
To prepare this trip, I diligently checked the weather forecasts two weeks prior to our vacation. However about one week prior, I almost cancelled our entire bookings, because the forecast basically said that it would have heavy storm throughout the entire week. I was disappointed and finally said:
“We are going, honey, just prepare to stay in the hotel for the entire time though…”
The first night after we arrived in Puerto Rico, I could not wait to turn on the TV and checked what the weather would be like the second day. I began to realize there were almost 10 different weather zones on the small island. The weather is not only different location from location, but also different hour from hour during the day. How could you possibly prepare for that?
In the past few years, I found myself spent lots of time fire fighting due to the inaccurate forecast. During this vacation, when I went from sunny beach to overcast hills and then got caught by a down pour at the rain forecast, I came to realize it is imperative to prepare for the worst all the time while wishing for the best (forecast accuracy). It made me think – when you manage your supply chain, is it more important to manage your forecast accuracy or to create a process to allow your supply chain respond to any abnormally of the forecast?
Through my Caribbean experience, I came to learn that to reduce the total lead time and lot size of your supply chain, create an early alert system to trigger actions and communicate changes openly throughout the entire supply chain are pragmatically more important than simply setting goals to create a 100% accurate forecast.
So how accurate does the forecast need to be in your mind?
As a finance person, I have hard time to picture planning without a fair amount of accurate forecast.
“accurate” does not mean “error-free”. Instead, it is a calling for serious commitment of effort and time to scenario analysis and simulation so that we have a fair assessment of what future may look like and the probability of each possible outcome—–we call this probability distribution.
Having an early alert system to trigger actions cannot happen without accurate forecast—since it should be designed aiming at the worse/worst scenarios in the probability distribution.
I am shocked by how many times supply chain people fire fight because sales folks do not have any forecast. As the direct revenue generator, sales need to provide first-hand information for forecast so that supply chain people can support. So that reason, my respect goes to supply chain people who have to do crsis management while little appreciation for their work.