Starting a Project

I tend to use an iterative process, with each iteration aimed at more detailed information than those before it. I like to start by assembling a group of user representatives and work with them to develop a single sentence that describes the project. I then work with them to expand that into a paragraph. Next, I work with them to develop a list of desired functionality. We then prioritize the items on that list as 1, 2 or 3. A number 1 priority is anything that is an absolute must have – without it, the project would be a failure. Number 2s are still very important, and 3s are nice to have. This is about the time I work with the group to come up with a name for the project. The final iteration of the high-level requirements gathering process is to develop high-level use cases and actors for the 1s and 2s. Each iteration in this process uses the output of the previous iterations to maintain a sense of scope. Anything that appears to fall outside the scope is discussed. It may be eliminated, or the scope may be changed. Typically, it gets eliminated. It’s less common to change the scope of the project. Once all of this is accomplished we can begin working on detailed use cases.