Estimate Cost and
Schedule Empirically
|
 |
|
|
| Roger Pressman | Capers Jones | Victor Basilli | Larry Putnam |
"A good cost estimator should never make an estimate that deviates from the actual cost at delivery by more than 20 percent. A good schedule estimate should be within 5 percent of actual schedule at least 95 percent of the time, and should not deviate by more than 12 percent less than the actual schedule."
--Capers Jones
Estimate Cost and Schedule Empirically
Software development projects historically have a poor record when it comes to project estimation. Statistics repeatedly show that the chance of a project completing on-time and under-budget is less than one in ten. Further, the chances are worse if the project is a large-scale, complex, software-intensive project, the kind the 16 Critical Software Practices were designed to help.
This estimation difficulty is not a reflection of our inability to forecast and analyze a new job. Rather, it is a reflection of the inherent complexity of software development projects, the multitude of ever-present factors that can influence cost and schedule, and the inadequacy of our estimation models to forecast the resources a project will take.
There are a number of ways to estimate the cost and schedule
for a software project. One of the most effective methods is
organizational experience-i.e., what the cost and schedule results
have been for similar efforts. The reason this estimation basis
can be so powerful is that it aggregates the multitude of factors
present in a given organization
Unfortunately, empirical cost and schedule estimation is often
more easily said than done. The "view from the trenches"
is that comparisons are hard. The rapid evolution of software
technology, languages, tools, and practices often makes it appear
that every software development effort is different. Often the
team composition and team dynamics will change. There is never
a perfect match between prior efforts and the one contemplated.
The cost and schedule estimation approach must be designed to take advantage of the information available about prior efforts to estimate future ones.
