In this video, we cover the naïve forecasting method. This forecasting method simply states that we forecast that this period will be the same as the previous period.
While Naïve forecasting can be useful in some situations, there are certain situations where this method of forecasting can be problematic. To demonstrate the naïve method, I’ve created a spreadsheet with 3 and a half periods worth of sales history.
Using the Naïve Method
The naïve method of forecasting dictates that we use the previous period to forecast for the next period. To demonstrate the pros and cons of this method I’ve created a % difference column. This column will show the % of variance between the Actual Sales column and the forecast. This will show you how accurate the forecast actually is. You can see the equation I used, =IF(Actual Sales=0,0(Naïve Forecast/Actual Sales)-1) , in the image below:
What this equation means is that if the forecast is less than the actual sales within that time period, then the % will be positive. A positive percentage means that what we actually sold is greater than what we forecast. A negative percentage means that we sold less than the forecast indicated we would sell.
To calculate a naïve forecast simple take the previous month of sales and plug it in next to the adjacent period. The equation for this method, =(Previous months actual sales) , is shown below:
Once you’ve applied the equation, you’ll notice that the equation has projected a positive percentage within 10%. That is a pretty good forecast relatively speaking, moving in the right direction. It is usually nicer to outperform your forecast promising less and selling more, but it can lead to some inventory planning problems that we’ll get into in more depth below.
For now, let’s apply the equation to the entire 3 years of sales periods we have to see what we get.
One of the very first things we can discern from this forecast is that we have periods of positive (sold more than we forecast) and negative (forecast more than we sold) periods.
These positive and negative periods are important to analyze because they potentially provide us with information about seasonality trends. These trends can help us to understand the performance of a product during particular times of the year.
Problems with the Naive Method
The next important thing to look at is the amount of difference. The higher the number, the worse it affects your company because a widely inaccurate forecast makes it impossible to plan your orders.
As you can see in the example above, Period 7 had a variance % of 41.92%. In most cases this would be good, we under-promised and over-delivered. However, it is possible that the company could have sold much more, but we only had enough stock to meet the demand we forecast for. Thus we see that the naïve forecasting method has issues.
Really the only situation that I have been able to see this method useful is when talking to individual sales people. They hit a number and then forecast that they’ll hit the same number next month.
While there are some situations that call for this method of forecasting, we recommend using a more robust approach. Avercast is a forecasting and demand planning software company. Avercast forecasting software is powered by 250+ forecasting algorithms to make supply and operations planning more accurate.
How it applies to inventory planning?
As we saw above, the naïve method has its benefits and its limitations. I want to preface this section by saying that as a supply planner managing inventory, or a demand planner planning for customer demand, in accompany you shouldn’t be using the naïve forecasting method. That being said, if your current forecast model is predicting demand at a worse rate than the naïve model, it’s time to change things up. The naïve method is the simplest form of business forecasting, but there are slightly more complex forecasting models that are almost as simple as naïve forecasting. We discuss some of these in other articles here on the blog. So, moral of the story? Naïve forecasting maybe a good basis initially, but once you have some real data and multiple products, move to something more robust.