[1]

B. Boehm, "Value-Based Software Engineering", ACM Software Engineering Notes, Vol. 28, No. 2, March 2003.

This paper provides a great introduction to Value-Based Software Engineering. It the seven key elements to VBSE (Benefits realization Analysis, Stakeholder Value Elicitation and Reconciliation, Business Case Analysis, Risk and Opportunity Management, Concurrent System and Software Engineering, Value-Based Monitoring and Control and Change as Opportunity). Each of these are described in detail and examples are given as to their usefulness.

Section 2 mentions model clashes which are certainly a big problem whenever groups with differing goals are introduced.

Section 3 of the paper describes steps for getting started towards VBSE.

In a section titled "Value-Based Monitoring and Control at the Organization Levl", the author mentions Dr. Basili's work on the Experience Factory and its Goal-Question Metric (GQM) approach. A "value-based version of the EF-GQM" approach is shown in in Figure 10. This is an idea that perhaps could be taken further but I'll want to read more on EF's.

[2]

B. Boehm and L. Huang. "Value-Based Software Engineering: Reinventing "Earned-Value" Monitoring and Control", ACM Software Engineering Notes, Vol. 28, No. 2, March 2003.

This paper essentially says the same thing as [3]. The same example for Sierra Mountainbikes is used.

[3]

B. Boehm and L. Huang. "Value-Based Software Engineering: A Case Study", IEEE Computer, March 2003, Vol. 36, No. 3.

This paper begins by describing earned-value systems, where a value is assigned for each task activity. At time T the project can be determined to be on track, or otherwise, based on how the cost of work performed differs from the cost of work scheduled.

The authors argue that earned-value systems make sense when business values change slowly but does not scale well in today's world of "tornado driven" changes in the IT marketplace. They mention that it is common to see projects delivered on time and near budget but where the business need has largely disappeared. Thus, earned-value systems become difficult to administer if the project's plan changes rapidly. They also mention that it neglects the actual value the project is earning for the organization.

Monitoring and controlling earned value starts with a project's business case. The authors use DMR Consulting Group's benefits realization approach and its results chain which links to assumptions. They use as an example an order entry system. They put together a schedule and a table of expected benefits. I assume there is some justification for these numbers but none is given. It should probably be assumed that these are the original numbers used to justify the program to begin with.

As for value-based monitoring and control, they basically state that they can use these numbers to direct the project as it progresses. Corrective action can be applied whenever (1) benefits are not realized, (2) prior assumptions become invalid, or (3) new and better opportunities emerge.