Skill Level Appropriate for all
- Forecasting is a fancy name for guessing about what you expect to happen in the future. Every business depends on forecasts to plan ahead and make critical decisions, so understanding how forecasting works is a really valuable skill. In this video we'll look at some key concepts that will help you use forecasts more effectively. We'll look at the two different types of forecasts, two common patterns that occur, and two different ways to measure forecast accuracy.
The first thing to remember about forecasts is that there are only two ways to create them. With a quantitative forecast you assume that the future will resemble the past, so you use statistics from historical data to try to predict the future. For example, if you've sold an average of 10 units per month in the past, you'll probably sell about 10 units per month in the future. While quantitative forecasts are built on historical data, qualitative forecasts are based on judgment.
For example, what if you're introducing a brand new product? You won't have any historical data to work with so you'll need to use your judgment to create a qualitative forecast. Qualitative forecasts can be based on the judgment of one expert or on the judgment of a group of people like a focus group. The next thing to remember about forecasts is that they often reveal patterns. For example, if your forecast goes up or down by the same amount each month, then that indicates a trend.
The next pattern to look for is called cyclicality or seasonality. For example, this is when you'd except a forecast for sales of mittens to be high in the fall and winter and low in the spring and summer. Seasonality is a repeating pattern, and it looks like waves in the forecast. It's possible to have a trend occurring at the same time that you have seasonality. The third thing you should know about forecasts is how to tell whether they're good or not.
There are two things to look for when evaluating your forecast accuracy, error and bias. The difference between your forecast and the actual results is called the error. You can expect to have some error in any forecasting process. Naturally, the less error you have, the more accurate your forecast will be. Now if your forecast is consistently higher or consistently lower than your actual results, it's biased.
A forecast that's biased too high, like this one, will lead people to be overly optimistic. A forecast that's biased too low can leave people unprepared for the work they need to do. While the basic concepts of forecasting are pretty straightforward, actually doing the math can be tougher. The good news is that spreadsheets like Microsoft Excel have lots of forecasting features built in. For a deeper dive, check out Wayne Winston's course, Excel Data Analysis: Forecasting.
Understanding how to create a forecast and how to tell whether your forecasts are accurate can help you do a better job of planning for the future, and better planning is guaranteed to make you more successful when managing your supply chain operations.