Measuring project value requires estimating what an organization is willing to pay in order to obtain the changes that the project will produce. At the project level a PPM system must measure the value contributed by a project and value that is given up if the project is delayed. The fundamental decision is what to do in the current year and what to delay.
Project value must be measured in terms of contribution to company objectives. Measuring project value requires the application of a rigorous value measurement method. The validity of the resulting measures requires that the measures satisfy specific conditions. Consider the following:
- In electric systems projects solve lots of different kinds of problems - that is they change things in the physical world (e.g., safety, reliability, customer satisfaction, and so on.). The trick is to apply a method that converts these changes into measures of value in the sense that the measures reflect how strongly the organization prefers to obtain the changes
- One characteristic of valid measures of value is such measures imply specific tradeoffs. For example if you have measures of value for two projects, A and B, these two measures of value allow you to explicitly state how much you would be willing to give up of project A in order to have project B. If your measures cannot be used to make this quantitative statement they do not reflect the value of the projects.
- However if the measures can be used to make this statement you can go one step further — you can directly convert your measures into the dollar value of projects. The only systematic method that I know of that produces measures of value with this characteristic is Multi Attribute Utility Analysis (MUA). For those that have a project portfolio system in place, there are a couple of simple questions that you can use to explore the validity of your value measurements.
- Pick a project that can be easily split into two pieces. For each piece develop a measure of value. Then ask the question is the sum of the measures of value for the two pieces equal to the measure of value for the total project? If not your system is measuring something but not value. This test was suggested by a colleague Lee Merkhofer.
- I used this test when I was hired by a company to evaluate their project ranking system. Senior management had hired a consulting company to develop and implement a system that would produce a ranking of projects. I simply ask the engineers in the company if, in their opinion, the ranked list provided credible results that is are the rankings consistent with how they would rank the projects. The engineers responded that the system did in fact produce a ranked list but, in their opinion, the ordering of projects was not consistent with how they would rank the projects. If you apply this test and get this result, at a minimum you have to question the validity of your value measurement system.