What is a Tracking Signal?

A Tracking Signal (TS) is a measurement used in forecasting and supply chain management to monitor the persistence of forecast error. It helps inventory managers and data analysts determine if a forecasting model is systematically over-predicting or under-predicting demand. Unlike absolute error metrics, the tracking signal indicates the direction of bias, making it a vital tool for maintaining forecast accuracy over time.

The Tracking Signal Formula

The calculation of the tracking signal involves two primary components: the Running Sum of Forecast Errors (RSFE) and the Mean Absolute Deviation (MAD). The formula is expressed as:

Tracking Signal = RSFE / MAD

Where RSFE is the cumulative sum of actual demand minus forecasted demand for each period, and MAD is the average of the absolute values of those forecast errors. By dividing the total bias (RSFE) by the average magnitude of the error (MAD), we get a ratio that tells us how many standard deviations the forecast is drifting from the actual performance.

How to Use This Calculator

Using our Tracking Signal Calculator is straightforward. First, calculate your RSFE by summing up all forecast errors (Actual - Forecast) for the observed periods. Note that this can be a negative number if you are consistently over-forecasting. Next, calculate your MAD, which is always a positive number representing the average absolute error. Enter these two values into the input fields above and click calculate. The tool will instantly provide the tracking signal and let you know if your model is within typical control limits.

Interpreting the Results

In most professional supply chain environments, a tracking signal between -4 and +4 is considered acceptable or "In Control." If the signal exceeds +4, it indicates that the actual demand is consistently higher than the forecast (under-forecasting). If the signal drops below -4, it suggests that the actual demand is consistently lower than the forecast (over-forecasting). When the signal moves outside these boundaries, it is a clear indicator that the forecasting model needs adjustment or that a fundamental change in market demand has occurred.

Frequently Asked Questions

Why is RSFE important? RSFE measures the total bias. Without it, you wouldn't know if your errors are canceling each other out or accumulating in one direction.

Is a tracking signal of 0 perfect? Yes, a tracking signal of 0 means that the cumulative forecast errors perfectly balance out, indicating no systematic bias in the model.