A project will see an increased likelihood of meeting objectives by integrating on-going consideration of uncertainty and risk throughout the lifecycle. Taking a lifecycle approach that is not just a pessimistic threat-based approach to risk management is recommended. Shifting attention from simple threat-probability-impact analysis to a broader view including uncertainty and variability, and examining sources in addition to potential events has advantages for evaluating complex interactions.
Examples include interfaces, interactions between stakeholders, organizations, functions, and processes. In every project we are faced with limited knowledge, unknown future decisions, changing influences, and multiple actors. Difficult situations with ambiguity, lack of clarity, lack of structure, unknown stakeholder bias still need to be managed successfully. For example, requirements are different from what a stakeholder may be ultimately interested in, and must be understand early.
- There are many actors in every project who contribute to requirements, communication, estimates, priorities, skills, etc. (Have we identified all?)
- Estimation uncertainty (bias and assumptions)
- Roles and responsibilities are a large part of any project leadership responsibility. Clear and measurable accountability for each position is essential.
- Uncertainties have a range of possible outcomes
- Complexity may contribute to uncertainty
An important project management technique to reduce risk is to start with clear objectives and understanding of the relative priorities of time and cost or other factors.
- Do all stakeholders, contractors, governance, other actors understand in detail their role and commitment associated with the project?
- If there is ambiguity or incomplete definition of the project charter (an initiative may require further refinement over time), a formal mechanism to review and evaluate value at milestones is important (e.g. Spiral Method).
Project Process Risk Analysis:
Examine project processes to identify:
- Assumptions that may no longer be valid-avoid the trap of always being one assumption behind.
- Testing a system based on the design vs. meeting original requirements-typically occurs when the test plan is developed late in the project life cycle leading to a focus on the function of the finished product only.
- Design risks that have been identified but not handled effectively. The consequences can be disastrous.
Over Confidence Risk: Mars Climate Orbiter Failure
A misunderstanding on the use of Metric vs. Imperial units in one file caused the spacecraft to crash into the Mars atmosphere and was lost.
One can imagine a working environment where JPL had been building and flying spacecraft for many years and began to believe that some of the work was well understood and not in need of close inspection, scrutiny, or testing despite tremendous change in spacecraft technology and mission complexity over time.
Risks could have been discounted as the relative level of organizational confidence in spacecraft navigation may have been fairly high. This could have led the various functions to limit time devoted to detailed analysis when faced with other demands, and executive leadership could have been shining the bright light of visibility on other areas. It is also interesting there was no monitoring that might have detected the small cumulative errors over the course of the nine month journey; perhaps another indication of confidence clouding the organizations ability to recognize and effectively address problems.