To design an effective RMP for a particular application, experience suggests that a limited top-down strategic view of project uncertainty is a sensible starting place. Sources of uncertainty internal to the project and corporate sources of uncertainty can be related and sized to provide an overview that is useful in its own right, as well as providing a basis for further more detailed analysis. Simple forms of quantitative approaches are needed, as illustrated later in Example 11.3. Such approaches must recognize that some sources of uncertainty are best left as conditions (assumptions), but semiquantitative risk-ranking approaches like that proposed by Baccarini and Archer (2001) are to be avoided, for reasons comparable with our case against 'qualitative' approaches to estimation and evaluation in Chapters 10 and 15. Uncertainty associated with the obvious lack of a structured bottom-up analysis basis means the tendency to optimism and underestimation of uncertainty discussed in Chapter 10 needs to be neutralized as far as possible, overstating uncertainty when in doubt.
This kind of initial top-down uncertainty appreciation may reveal levels and forms of uncertainty some people do not want to address. However, from a risk management perspective, it really is a waste of everybody's time to do otherwise. If bearing this kind of bad news involves 'constructive insubordination' that is likely to be so unwelcome that it may cut short the project risk analyst's career in this organization, he or she may wish to keep this part of the analysis to themselves for the time being; but, helping to keep bodies buried for any length of time is inappropriate.
Example 6.3 illustrates that an important reason for undertaking a top-down view of uncertainty is to determine where the limits of the project manager's responsibilities for managing project-related uncertainty lie. The rest of Part II assumes a simple division of 'external' and 'internal' sources of uncertainty between board level and project management for most illustrative purposes, but a more complex, hierarchical division of risks can be helpful, for reasons considered in Chapter 9.
Was this article helpful?