What’s an inventory forecast?
Inventory forecasts will help you estimate the future demand for your products in a short, mid, or long-term period and it’s a fundamental part of any successful inventory planning operation.
Knowing those forecasts in advance will allow you to improve customer service, inventory control, and capacity management.
We could forecast the cost figures in an inventory control model; these are forecasts of the costs that will apply in the future when it is time to place an order. Another area to forecast is the lead time, which is generally getting shorter. Whatever we decide to forecast, remember these basic principles:
- The forecast will be more accurate for the short term. The farther you go, the less accurate you are.
- The forecast will be wrong. There will be a forecast error, and it is essential to know that percentage.
- The forecast is no substitute for actual demand.
Type of forecasts
There are so many ways of forecasting, and different things to forecast, that no single forecasting method is always the best. We have to choose a method that suits our needs. In either case, we have to choose one based on time:
- Long-term forecasts look ahead several years
- Medium-term forecasts look ahead between three months and a year
- Short-term forecasts look ahead to the next few weeks
There are several ways to base the forecast. They could be based on historical data or judgmental estimates. The forecasting method utilized will depend on the data sources available. If there is a reliable demand history of a product, it can be the forecasting source.
Forecasting techniques
As I’ve stated before, there are several techniques to calculate forecasts. We’ve already covered the economic order quantity technique in great detail in the past, which will determine how many products you need to buy at any given time. This post will briefly analyze the most used ones that use historical data.
The three methods described below will use the same historical data from the past three months (October, November, and December 2022) to forecast the next three in 2023.
Let’s see the first method…
Moving average
The moving average method it’s a simple average calculation. This technique averages a user-specified number of months to project the next month’s demand.
| October 2022 | 114 |
| November 2022 | 119 |
| December 2022 | 137 |
| January 2023 | ? 114 + 119 + 137 = 370, 370 / 3 = 123 |
| February 2023 | ? 119 + 137 + 123 = 379, 379 / 3 = 126 |
| March 2023 | ? 137 + 123 + 126 = 379, 386 / 3 = 129 |
Weighted moving average
The weighted moving average lets you assign weights to the historical data. This method works better for short-range forecasts of mature products.
Note: The assigned weights must total 1.00, with the most recent data receiving the greatest weight. In this example, let’s assume the weights are: 0.6, 03, and 0.1
| October 2022 | 114 |
| November 2022 | 119 |
| December 2022 | 137 |
| January 2023 | ? 137 * 0.6 + 119 * 0.3 + 114 * 0.1 = 129 |
| February 2023 | ? 129 * 0.6 + 137 * 0.3 + 119 * 0.1 = 131 |
| March 2023 | ? 131 * 0.6 + 129 * 0.3 + 137 * 0.1 = 131 |
Linear smoothing
This method is similar to the weighted moving average. However, a formula is used instead of assigning weights manually, and now the weights decline linearly. This method works better for short-range forecasts of mature products.
Weight for one period prior = 3/(n^2 + n)/2 = 3/(3^2 + 3)/2 = 3/6 = 0.5
Weight for two periods prior = 2/(n^2 + n)/2 = 2/(3^2 + 3)/2 = 2/6 = 0.3
Weight for three periods prior = 1/(n^2 + n)/2 = 1/(3^2 + 3)/2 = 1/6 = 0.16
| October 2022 | 114 |
| November 2022 | 119 |
| December 2022 | 137 |
| January 2023 | ? 137 * 0.5 + 119 * 0.3 + 114 * 0.16 = 127 |
| February 2023 | ? 127 * 0.5 + 137 * 0.3 * 119 * 0.16 = 129 |
| March 2023 | ? 129 * 0.5 + 127 * 0.3 * 137 * 0.16 = 130 |
Conclusion
Forecasts are based on historical data and models. They are not perfect; they will contain errors that might be caused by models, data, or interpreting the results, but a strong estimate is better than no estimate.
In future posts, we will look deeper into more advanced forecasting methods (like exponential smoothing, the least square regression, and others), analyze their pros and cons, and how they can be automatically generated using inventory management systems.