Extrapolating current or known performance to predict future performance over a period of time.

Example: Assume a company wants to estimate how many defects will be produced for the entire year. They’ve had a total of 167 defective parts over the first four months of the year (Jan-Apr). This equates to an average of approximately 42 (41.75) defects per month. Applying this run rate of 42 to each of the remaining eight months, it equals 336 in total. Adding 336 + 167 gives 503 total defects for the year.

This is a simple example to show how run rate works. In reality, to get a more accurate prediction, a run rate is used in conjunction with other prediction data, which could include historical trends, seasonal trends, known events, 12-month run rate, rolling rates or performance trending.