What it means
When running a business, you constantly need to predict future outcomes, such as next month sales, quarterly cash flow, or annual staffing needs. Once the actual period passes, you want to know how close your guess was.
Mean Absolute Percentage Error, often called MAPE, makes this easy to calculate and understand because it translates forecasting errors into percentages rather than confusing dollar amounts or units. To find this metric, you look at the gap between your forecast and the actual result for several items, turn those gaps into percentages of the actual numbers, and then find the average.
For instance, if you predicted selling 90 units and sold 100, the gap is 10 units and the error is 10% of the actual result. By doing this across all your products or months, you get a single score that shows your overall forecasting reliability.
This matters because bad forecasts lead to wasted money, either by holding too much stock or running out of critical items. Business leaders use this metric to track whether their forecasting methods are improving over time.
A finance team might set a target, such as keeping the error rate below 10%, to ensure budgets and purchasing plans remain dependable and realistic. Imagine throwing darts at a board.
If you miss the bullseye by two inches every time, you are consistently off. This metric measures the average size of those misses as a percentage of the total distance to the target, showing how good your aim is overall.
In practice
Real-world examples
Example 1
A boutique coffee shop owner forecasts selling 1,000 lattes in June, but actually sells 1,200. The error is 200 / 1,200 = 16.7%. If July has an error of 10%, the MAPE across the two months is (16.7 + 10) / 2 = about 13.4%, a figure the owner uses to judge overall forecasting accuracy.
Example 2
A regional courier firm predicts it will deliver 5,000 packages on Tuesday, but handles 4,500 due to bad weather. The error is 500 / 4,500, which is 11.1%. Tracking this helps managers schedule the right number of drivers for future shifts.
Example 3
An online clothing retailer forecasts 2,000 winter coat sales in November, but sells 2,500 following a sudden cold snap. The error is 500 / 2,500 = 20%, showing purchasing managers where seasonal demand spikes catch them off guard.
Think of it
Imagine throwing darts at a board. If you miss the bullseye by two inches every time, you are consistently off. This metric measures the average size of those misses as a percentage of the total distance to the target, showing how good your aim is overall.
Formula
Calculation
MAPE = (1 / n) x the sum of [ |Actual - Forecast| / Actual ] x 100, where n is the number of periods and the vertical bars mean you ignore whether the error is positive or negative. Do this for all periods, add up the percentages, and divide by the number of periods.
Single period example: actual sales = 200 units, forecast sales = 180 units. The difference is 20 units and the percentage error is 20 / 200 = 0.10, or 10%. If you have two months with errors of 10% and 20%, the MAPE is (10 + 20) / 2 = 15%.
Dollar example over three months: actual revenue is $100,000, $80,000 and $120,000, against forecasts of $90,000, $88,000 and $108,000. The errors are $10,000 / $100,000 = 10%, $8,000 / $80,000 = 10% and $12,000 / $120,000 = 10%. The MAPE is (10% + 10% + 10%) / 3 = 10%.Case study
Seen in the real world
This is a fictional, illustrative scenario. Oakwood Furniture, a growing manufacturer of dining tables, struggled with inventory costs because their sales forecasts were unreliable. The management team decided to track their forecasting accuracy using Mean Absolute Percentage Error to improve their production planning. Over the past four quarters, Oakwood predicted sales of 500, 600, 550, and 700 tables.
The actual sales figures turned out to be 400, 650, 500, and 630 tables. The absolute percentage errors for each quarter were calculated as 100 / 400 = 25%, 50 / 650 = 7.7%, 50 / 500 = 10%, and 70 / 630 = 11.1% respectively. Averaging these figures gave Oakwood a baseline MAPE of (25 + 7.7 + 10 + 11.1) / 4 = 13.5%. The finance director shared this data with the sales and operations teams, setting a goal to bring the error rate below 10% by the end of the year.
By factoring in recent market trends and historical seasonality, the team refined their approach. In the following year, their error rate dropped to 8.5%, which significantly reduced excess timber storage costs and improved cash flow. The finance director also noticed that the first quarter, at 25%, drove most of the original error, and she asked sales to explain what had been missed, such as a delayed customer order.
Watch out
Common mistakes
- Using this metric when actual sales figures can be zero, which breaks the math because you cannot divide by zero.
- Treating a high percentage error on a very small product line the same as a high error on your core revenue generator.
- Failing to look at whether you consistently overforecast or underforecast, since the absolute values hide the direction of the error.
Questions
People also ask
What is considered a good MAPE score?
Generally, an error rate under 10 percent is considered very good, while anything under 20 percent is acceptable for most business planning purposes.
Why use percentages instead of actual numbers?
Percentages let you compare accuracy across different products, departments, or time periods on an equal footing, regardless of their scale.
Can this metric be used for financial budgeting?
Yes, it is widely used to test how accurately financial teams predict revenues, expenses, and cash flow against actual results.