Behavioral Earned Value

Behavioral Earned Value (BEV)

Behavioral Earned Value is the application of behavioral and neuroscience to earned value methods, and uses the capabilities in measurement that earned value provides to see issues such as consistent optimistic (or pessimistic) planning and forecasting, giving us an opportunity to investigate further for the presence of relevant biases and other decision errors that cause poor project prediction and performance.

Traditional technical-only Earned Value (EV) gives us the ability to measure the performance of a project. It takes a predicted value (usually from the baseline plan) and compares the actual performance to what the performance was planned to be. 

But Earned Value is an under-appreciated method that can give us insights into behavior that hadn't been considered previously. The issue with many behavioral phenomena is that sometimes measurement of it is hard to accomplish. EV helps solve some of this problem. 



...Being able to know how accurately a PM, Estimator, or other forecaster tends to predict their work. And once their prediction errors are known, they can be corrected, which translates into more accurate baseline planning, more reliable monthly forecasts, lower schedule and cost variances, and ultimately higher customer trust.


Advantages of BEV

Traditional EV only measures performance against the plan. With traditional EV, however, we can still measure how well a PM or Estimator predicted (planned) the scope of work. 

With advanced EV methods we can take it a step further. By taking the next month's forecast (BCWPf) and treating it like a baseline, we can measure performance against the forecast and see how PMs and other forecasting personnel are performing to their forecasts each month, giving us an average forecast rate. Because humans tend to forecast optimistically due to the Planning Fallacy, we can begin to measure these averages by forecaster, project, or organization and start to make corrections through training, coaching, and choice architecture. 

The advantage is that with BCWPf (or Forecast P) we can begin measuring prediction error month to month and start making incremental adjustments, instead of waiting for baseline performance to come through months or years later. This average prediction error can then be improved and monitored for favorable change. Then your organization can place its most reliable forecasters in the war room for baseline planning, thus improving the overall reliability of your planning efforts.


How is it Done?

Here's one example. Look at Activities A and E. Note that each has difference between what was predicted and what the actual performance was. Note that Activity A can typically be measured with traditional EV methods. However, many times we tend to blame this performance on those performing execution of the project, not those who planned it. 

Now let's look at Activity E. Note that this activity is right of the data date, and still hasn't occurred. It's in a predicted or forecasted state. However, note that if previous forecast performance for the forecasting personnel tends to be 40% optimistic, we can apply an assumption of how we expect the activity to actually perform. 

Let's take it a step further. If we had measured Forecast P and its performance for the past 3 or 4 months we would have an average optimistic (or pessimistic) rate for the forecasting personnel over those periods (this is where we can actually start to pull the strings on what's going on and where we can make improvements). To measure the monthly forecast against the performance we take the earned value minus the forecasted value, and then divide this number by the earned value, giving us a Behavioral Forecast Rate (BFR). It might look something like this: (BCWP-BCWPf)/BCWP = BFR, or (12,850-18,000)/12,850 = -0.40078.


The Behavioral Part

Now that we have a consistent average measurement, we can look for potential causes by individual, project, organization, department, etc. Some behavioral concepts that could then be considered might be optimism bias, overconfidence effect, deliberate ignorance, information avoidance, courtesy bias, or authority bias. Once mitigations have been applied, we can then remeasure on a monthly basis to see if there are improvements and start to improve prediction accuracy in both planning and monthly forecasting. Because of the underlying inherent nature and cause of some of these behavioral issues, the project may also see other positive residual outcomes from using these methods, such as reduced risk realization and greater project safety.





Join your peers and become a member of the most advanced project management endeavor, the building of #projectscience through the neuro, behavioral, and cognitive sciences! Behavioral Economics has made great strides, so what are we waiting for?


Be the first to comment

Please check your e-mail for a link to activate your account.