At this point in the process, the software
has been defined, developed, tested and fielded. This step is the
"post-mortem" step. That is, did the development of the
software achieve the goals that were set out for it.
If the software did not meet the goals, why not and what lessons can
be learned from it? If the software did meet the goals, can procedures
and processes be applied to other projects underway to aid them?
Finally, this step provides a benchmark that new projects can be
measured against. That is, it should be easier to gauge how long a new
project should take if it is similar to a project that has already been
done.
In our experience, this step is the one that is most often overlooked
as the software is already completed, in use by the users and has met
(ideally) all the goals defined for it. We believe that if the
evaluation step is performed, however, future projects can be completed
with much more confidence in the budgeting and timeline to completion.