Improving project forecast accuracy by integrating earned value management with exponential smoothing and reference class forecasting

- Integration of a single smoothing parameter in known EVM framework ensures easy use
- Integrating exponential smoothing allows forecasts based on more recent performance
- Able to consider effect of natural performance improvement and corrective actions
- Smoothing parameter can be dynamically adjusted according to management insights
- Huge potential increase in time and cost forecasting accuracy, especially with RCF

In this paper, the earned value management (EVM) project control methodology is integrated with the exponential smoothing forecasting approach. This results in an extension of the known EVM and earned schedule (ES) cost and time forecasting formulas. A clear correspondence between the established approaches and the newly introduced method - called the XSM - is identified, which could facilitate future implementation. More specifically, only one smoothing parameter is needed to calculate the enhanced EVM performance factor. Moreover, this parameter can be dynamically adjusted during project progress based on information of past performance and/or anticipated management actions. Additionally, the reference class forecasting (RCF) technique can be incorporated into the XSM. Results from 23 real-life projects show that, for both time and cost forecasting, the XSM exhibits a considerable overall performance improvement with respect to the most accurate project forecasting methods identified by previous research, especially when incorporating the RCF concept.