Discovering Risk With 3 Point Analysis
3 point analysis is a method of estimating the time that a project feature will cost. Rather than define a single cost value, 3 points are chosen;
- Optimistic. The cost of implementing a feature if everything went perfectly.
- Likely. The most likely scenario, e.g. how long the feature will probably take.
- Pessimistic. The cost of the feature if everything goes horribly wrong.
Defining three cost values for each estimate doesn’t take a great deal more time, but they can give a project manager a fantastic insight in to the likely upper and lower bounds on the amount of time a project should take and where the risk lies is in the project.
Ordinarily, with a single cost per project item, a project manager will have the sum of the estimated costs in order to estimate the total cost of the project. While useful, that doesn’t give an accurate representation of risk. How much longer the project would take if things don’t go to plan is important. The sum of the optimistic costings is the best possible time that the project could take, while the sum of the pessimistic costings is the worst. Adding the optimistic and pessimistic cost estimates gives a deeper insight – a project that takes 10 times longer if things don’t go to plan is incredibly risky.
This analysis can help test the project as a whole, they’re also useful on individual requirements. If the difference between the optimistic and pessimistic costs is small then there is a consequently small risk – even if things go wrong the budget for implementation cost isn’t adversely challenged. Conversely however, if there is a large difference, then a lot can go wrong and the feature carries a great deal of risk to the project deadline and budget.
Usable Requirements has 3 point cost analysis for requirements rolling out in the next week. Every requirement can be estimated with optimistic, likely and pessimistic values. The project dashboard gives project managers a simple graphic view of both total costs and risk based on these estimates.