Step 1 – Define integrated project success criterion
With multiple success criteria decision making is complicated – increasing one of them we may decrease another. There is a need for some weighting factor that may be used for decision making. It is necessary to be able to measure overall benefits of projects and portfolios, to be able to compare options and to select the best management decisions. We suggest to set one integrated criterion of the project/portfolio success or failure.
One of the potential approaches is to use money for measurement of everything. For example defining the cost of one day for project acceleration and delay we will be able to estimate if it is profitable to pay more for faster performance and if project performance was successful if it was late but saved certain amount of money.
Step 2 - Create optimistic project schedule model
Optimistic model is based on optimistic estimates of all project parameters and includes only most probable (90% probability or larger) risk events. This model is used for setting performance targets for project workforce. It is clear that optimistic targets will not be achieved but in any case performance targets shall not include contingency reserves or they will be lost (Parkinson Law).
Step 3 - Simulate risks and set reliable targets for project management team
Project management team shall have time and cost buffers for managing project risks and uncertainties. Project or phase buffer is a difference between target value and the value for the same parameter in the optimistic schedule. Targets shall be set using risk simulation. These targets shall have reasonable probabilities to be met (usually in 70-80% probability range).
Project and phase targets and buffers may be created not only for integrated project success criterion but also for separate parameters like project cost and duration. Probabilities to meet project/phase targets are called success probabilities.
Step 4 - Set project sponsor targets
Management reserve for unknown unknowns is usually created basing on past performance data. When these data are missing or not reliable project sponsor targets are set using the same risk simulation model but with higher probability to be achieved (usually in 90-95% probability range).
So project has a set of targets – tight targets for project team, reasonable targets for project management team that include sufficient contingency reserves, and more comfortable targets for project sponsor that include additional management reserves.
Step 5 - Estimate buffer penetrations
It is natural that project will be late to optimistic schedule and project/phase buffers will be penetrated in the process of project execution. It is necessary to be able to estimate if these buffers are still sufficient and if project performance was better or worse than expected. The natural way for estimating buffer penetrations is calculation of current probabilities to meet the targets. If new probabilities are higher than initial project performance was better than expected though success probabilities depend not only on internal factors. If project performance was perfect but new risks were identified success probability may become lower because initial contingency reserves did not consider these new risks.
Step 6 - Analyze success probability trends
Current success probabilities show project status but project status information is not sufficient for decision making. Decision making shall be based on the analysis of project trends.
If the probability to meet project target is rising then project buffer was consumed slower than expected, in other case project buffer was consumed too fast and project success is endangered. Management decisions shall be based on the trend analysis. Even if current status is good (success probability is high) but the trend is negative corrective actions shall be considered.
Success probability trends are the best integrated performance indicators – they take into account project risks, they depend not only on performance results but also on the project environment changes.