Moving Average Forecast Calculator for Demand Planning
The simple moving average is the oldest baseline in demand planning: forecast the next period as the average of the last n periods. It is quick to explain, easy to audit and a good benchmark that fancier models should beat.
Your numbers
Shorter windows react faster, longer windows smooth more.
Your results
A simple moving average assumes demand is flat around a stable level. It lags a trend and ignores seasonality, so treat it as a baseline, not a finished plan.
Baseline forecasts for every item
Optional: we can sketch how to generate and score baseline statistical forecasts directly from your ERP demand history.
Optional. The tool and every export work without this. No spam.
How to use the calculator
Enter your most recent six periods of demand, oldest first, using the same unit and bucket (weeks or months). Choose a window of 3, 4 or 6 periods. The calculator returns the next period forecast, the naive forecast (just last period), the six period average and the percentage gap between the forecast and last period. Try each window and compare the results to see how sensitive the forecast is to the choice. Large differences mean the series is trending or volatile.
The formula and the window trade-off
For a window of n periods, the forecast is the sum of the last n actuals divided by n. A 3 period window gives each period a weight of one third, so it follows recent changes quickly but also chases noise. A 6 period window gives each period one sixth, so it is smoother but lags a real step change by roughly half the window length. There is no universal best window. Test candidates against history and keep the one with the lowest WMAPE and bias.
- Short window: fast reaction, more noise.
- Long window: stable, but slow to see a trend.
- Window equal to the seasonal cycle length removes seasonality but loses its shape.
Worked example
With demand of 120, 132, 118, 140, 135 and 148, the 3 period average is (140 + 135 + 148) / 3, which is 141. The 6 period average is about 132.2 and the naive forecast is 148. The series is rising, so the longer window sits well below the latest actual. If the rise is real, a trend-aware method such as Holt exponential smoothing will do better; if it was a one-off, the moving average is the safer choice.
Common mistakes
Do not apply a moving average to seasonal items without adjusting for the seasonal index first. Do not include periods distorted by stockouts, one-time orders or promotions without cleaning them, because the average carries that spike for n periods. Do not use demand that was constrained by capacity as if it were true demand. And remember an ERP forecast model labelled moving average may use a different window or weighting, so check parameters before comparing.
Frequently Asked Questions
How do you calculate a 3 period moving average forecast?
Add the last three periods of demand and divide by 3. That value becomes the forecast for the next period. When the next actual arrives, drop the oldest value, add the new one and repeat. The window slides forward one period at a time.
Is moving average better than exponential smoothing?
Neither is always better. A moving average weights the last n periods equally and ignores older data, while exponential smoothing weights recent data more and keeps a long memory. Exponential smoothing usually needs less data stored and adapts more smoothly, but moving average is easier to explain.
What window length should I choose?
Pick the window that minimises error on your own history, usually between 3 and 12 periods. Shorter windows suit volatile or changing demand, longer ones suit stable demand. For seasonal items, deseasonalise first, then apply the average.
Why does my moving average lag a trend?
Because it averages older, lower values with newer, higher ones. In a steady upward trend, a window of n periods lags by about (n - 1) / 2 periods. Use a trend method like Holt or add a trend adjustment if lag is costing you.
If you want baseline forecasts like this generated for every item in your ERP and compared automatically against your planner overrides, we can build that as a read-only service.
Related Tools
Exponential Smoothing Forecast Calculator (Simple and Holt)
Enter the prior level, trend, latest actual and smoothing constants to get the simple and Holt trend forecasts, with a chart showing how alpha changes the result.
Supply Chain & InventorySeasonal Index Calculator and Worksheet
Calculate the seasonal index for a month from three years of history and apply it to next year's annual forecast to get the monthly forecast.
Supply Chain & InventoryForecast Accuracy MAPE and WMAPE Worksheet
Fill in actual and forecast per SKU and get item-level percentage error, plus portfolio MAPE, WMAPE, bias and forecast accuracy. Exports to Excel.
Infor M3 & CloudSuiteM3 Exponential Smoothing Forecast Calculator (Holt Trend)
Step a Holt level and trend forecast forward one period and see how the smoothing constant changes the result.
More Supply Chain & Inventory tools
See all- Oracle Fusion Inventory Turns and Carrying Cost Calculator
- Distribution ERP Fill Rate Calculator
- Distributor Rebate Tracking Worksheet
- Branch Transfer Planning Worksheet
- Distributor Slow-Moving Inventory Review Worksheet
- Distributor Vendor Performance Scorecard Worksheet
- Distributor GMROI Calculator
- Distributor Backorder Cost and Fill Rate Calculator
- EOQ with Vendor Minimum Order Calculator