In a discussion on the german network XING on software engineering models and Tom DeMarco, I came to a simple structure on how a process model (or other “standards”) come to be and when or how they flourish or persih. I would like to propose a standard meta-process on process-model-life-cycles. It can be graphed as follows:
It contains 4 phases:
- Phase 1 – The Hatching
In this phase innovating companies build a set of processes that greatly increases their quality and ability to perform. They talk about this on several events and an interesting label is attached to it. Other innovative people are atracted to it and start to implement similar processes, always adapting them to their need. Initial documents are generated and a simple ownership body is set up.
- Phase 2 – Adolescence
As more people learn about the new standard, it undergoes a phase of quick growth. Innovative consultants jump onto the wagon and drive expansion. Initial software vendors create products supporting the young model. Not only adoption rate increases, but also the usefulness is increased during this phase, since it incorporates links to other standards, improves the quality of documentation, etc.
- Phase 3 – Shism
Right now the model has reached high acceptance within the relevant communities it seen more and more as a gold mine. Consultants pile upon the term and try to sell anything under its name. Peoples résumés are rewritten indicating they have been implementing the process model for several decades (usually longer than the model is in existance). Software vendors only remotly associated with the process model start selling their tools as the out-of-the-box solution for it. The people that created and have driven the process model up to this point start to turn their attention to new things. And now the shism occurs, the process model is either doomed for stardom and is granted an own ISO standard, or it returns to the crypt of gartners buzz words.
- Phase 4 – Old Age
The standards still increase their adoption, but much slower than it did before. New revisions try to expand the scope of the standard in a feeble attempt to gain more market share. This is true for both types of process models, the starred standard as well as the soon to be forgotten one. They both generate less and less benefit to companies embracing them. Some will linger on for quite some time, others will die more quickly.
Does this remind you of something?