Project Performance Patterns

EVM systems forecast final project outcomes based on current & assumed future performance, & trigger early warning signals when baseline outcomes are exceeded.

Earned Value Management metrics PV, AC and EV and ES extension, measure cost/schedule performance of a project in progress at a point in time (i.e. actual time AT), which reflects the amount of work delivered to date. Similarly, by analysing remaining work and making assumptions about future performance, it’s possible to forecast a project’s final outcome.

Estimate to Complete

From analysing the remaining work, two ways exist to develop the cost/schedule Estimate to Complete (ETC),

  1. Management ETC, which is developed by workers and/or managers and is most commonly used.
  2. Calculated ETC, which is developed using EVM/ES metrics.

Rather than relying on gut feel to predict project outcomes, calculated ETC provides a more thorough approach.

Performance Factor Patterns

Predicting final cost/schedule project outcomes based on assumptions about future performance for remaining work is challenging. Not only do project characteristics need to be considered but also the consideration of whether past performance will affect future performance needs to be evaluated. Consequently, three performance patterns exist,

  • Future performance will be according to plan. Past performance is not a good predictor of future performance. Project problems or opportunities of the past will not affect the future, and remaining work is expected to be done according to the planned work of the baseline schedule. In this situation, deviations from the expected time/cost performance as predicted in the baseline schedule are assumed to be one-time deviations and have no impact on the remaining work. In this case, the performance factor (PF) is equal to 1.
  • Future performance will be according to current time “or” cost performance. Past performance is a good predictor of future performance. In this case, problems or opportunities of the past will affect future performance. In this case, problems or opportunities of the past will affect future performance, and the remaining work will be corrected for the observed efficiencies or inefficiencies of the past. In this scenario, the performance factor is equal to SPI or SPI(t) for time forecasting and equal to CPI for cost forecasting.
  • Future performance will be according to current time “and” cost performance. Past performance is a good predictor of future performance. In this case, problems or opportunities of the past will affect future performance. In this case, cost and schedule management are inseparable, and hence the current performance is measured by a combined time and cost performance. In this case, the performance factor is equal to the Schedule Cost Index defined as SCI = SPI * CPI or as SCI(t) = SPI(t) * CPI or as a slightly modified version of these SCI or SCI(t) formulas.

Forecast

The time forecast of a project at completion equals,

  • Actual Duration of Work Delivered + (Planned Duration of Work Remaining / Performance Factor)

The cost forecast of a project at completion equals,

  • Actual Cost of Work Delivered + (Planned Cost of Work Remaining / Performance Factor)

Summary

Intuitively, knowing how quickly a project has progressed and how much it has cost to date, allows consideration of when a project will finish and how much it will cost. While in general, there is nothing wrong with management determined forecasts, they do tend over time to be inaccurate and unstable, which compromises management decision making. Instead, EVM calculated forecasts are more accurate and stable, and better supports actionable data-driven insights, especially when future performance is adjusted for one of the three performance patterns.

Learn More

If you would like to know more about leveraging data-driven actionable insights for your project schedule, then feel free to contact me on itierney@pminsight.com.au.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.