Each project has a
repository, and each project collects these data to run their own projects. At Level
3, these metrics are bubbled up from the projects and studied cumulatively to try
to predict trends across the projects. For example, if one project was late getting out
of the Testing phase because many errors were found, did that project skimp on
the Requirements phase? Is there a correlation? If all of the project data are studied
from all of the projects, and all of the projects except one had the same problem,
what is it about that one project that made it different? Was it better? Can we
devise a ???standard??? length of time to be set for requirements to improve functioning
downstream (like in the Test phase)?
Once an organizational-level repository has been built (based on historical,
project-level data from the past) any project-level repository can also use the
organizational-level repository as its foundation for building or updating its own
(project) repository. This gives the project manager an idea of how to plan his project,
where the bottlenecks commonly occur (or occurred in the past), and how long
to schedule activities.
Pages:
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255