Can the not-terminated, iterative development of a software application be seen as a project?

Question

I'm thinking of whether it is possible trying to apply project management methods and frameworks, e.g. PMBOK, PRINCE2 or something like RUP for software applications, that are still under active (iterative) development, but do not have a specified ending . As I've learned, a project is always a temporary endeavor, so something that is not terminated is not a project. What do you think?

In project management a project consists of a temporary endeavor undertaken to create a unique product, service or result. (Project Management Institute. A Guide to the Project Management Body of Knowledge (PMBOK Guide), Third Edition.)

Answer 1

If it is not a temporary endeavor, I think the term "project" does not fit well.

But you did not only ask if the term "project" applies, but also "can these project management methods and frameworks be applied?". Sure, you can probably try to use a full-blown project management approach for iterative software development by interpreting each cycle as a mini-project. However, this is IMHO somewhat the wrong question, because it gives you a misleading answer.

The real question you should ask is: does this make any sense? Or are you just shoehorning a situation into some kind of management method which does not really fit?

What you describe is what I would call "product development", as opposed to "project development". To my experience, product development, especially when done with a small team and in cycles of length of 3 weeks at maximum, simply does not require lots of the heavyweight stuff from the project management methods you mentioned. Especially anything like "several hundred pages requirements specs", "a monthly status reports of a dozen pages" or "tracking progress using large GANTT diagrams" are mostly obsolete if a product is evolved in small cycles.

That does not mean there is never a need for "project management" in product development. Especially when you start with a new product, or when you plan a new "major release" or a new module for the product with a given specification, defined goals and a fixed time frame, then this can require classical project management methods. However, at least when such a new version or module is "finished", and maintenance as well as short-cycle evolvement starts, these methods are seldom helpful, and agile methods like Scrum will be much better suited.

Answer 2

Yes. Unless they've changed it since I last studied in about 2004, the PMBOK definition of a project does require the endeavor to be finite in duration with solid objectives and scope. That being said, I think that modern software projects can be treated and managed effectively as a "project". There is nothing in the PMBOK definition of a project that says there can never be a follow on project to extend a previous project with new objectives and scope, and to achieve that by re-using the output of another.

The continuous development/iterative "project" still should be structured with finite goals and timeline of each iteration. But this type of project doesn't need the ramp-up time or end transition time you would expect on a project that is not picking up where another left off so there are economies gained.