Measuring risks, finding the correct metric(s)

Splitting time is a form of art

“What is a cynic? A man who knows the price of everything and the value of nothing.”

Mutual trust is a wonderful target, but it is a difficult standard to be maintained during all project’s time. At a lower level it is necessary to share some easy to be used as metric for evaluating the outcome through the acceptance of the single outputs .

Every stakeholder has his/her own target to pursue. Project’s outcome can be viewed as the “invisible hand” that pushes organized people together for reaping some future benefits.

Money & Time are elements that form “a necessary but not sufficient condition”

Quality, intended as passing the established threshold of successful tests, can complete the definition of metric. These are the concepts sustaining it:

1. Involving stakeholders in each (broad) category:
   1. Customers as:
      1. Suppliers of “real operational” conditions (i.e. “Boundary values”)
      2. Beholders of “final” value, either money (price), time, or both. It has to be extended to evaluation of each capacity and feature.
2. Producers as:
   1. User of chosen product (e.g. programming language, architecture, COTS)
   2. Acceptor of the project’s major conditions (“Iron Triangle”) to become accountable for their estimations that must include tolerances based on a proper distilled history (more to come on Bayesian analysis – i.e. subjective experience).
3. Easy to be framed into a commercial contract.
4. Using passed tests as metric, produces figures. This can be used for improving the productivity’s analysis.
5. They offer essential elements for Risk Management, both in terms of “sensible areas” and sources for working out the correct “likelihood”.
6. “Decision points” can be considered essential forms of testing without invalidate the meaning of it. In this case, either outputs or just concepts are evaluated (tested) in a specific environment.

Limits and caveats

IT projects cannot be limited to software development. Nor adding figures can be a reasonable definition for accountancy.

This introduces the first and, in my opinion, huge limit to the above presented statement.

Some operations are not easy to be tested in a “scientific” (repeatable) way (e.g. documentation, most of the business analysis, management). Others, like complex architectural operations (e.g. DB) require a lot of energies (time and money) to be seriously evaluated with long term strategies.

The correctness, if not the same meaning, of tests depends upon stakeholders’ collaboration. This requires overcoming the sense of secrecy about their needs and way of working.

Conclusion

The concept of “done” is viewed as watershed between producers and managers. Then, among various categories of stakeholders (from approvers to final users).
Once this big mountain is crossed, with the help of collaboration and mutual understanding, there is another challenge:
Producing and managing estimations.

More readings

Martin Fowler offers an excellent lesson about how communication can make shape leadership.

Image courtesy of:  Science Museum (London)