Model with purpose
There are many reasons why you'd process model. Those reasons may range from the requirement of a visualization to process improvement. Interestingly enough, there was a survey conducted recently by the good folks at BPTrends looked at all possible aspects that managers, architects and the like of people in our profession do who conduct any kind of process modeling and why they would do such a thing.
It was clear to see to see that there were two winners in that space.
- People process model to communicate, and,
- People process model to improve processes.
What can we learn from that? One thing's for sure. You need to be able to articulate what your purpose is. Use those two aspects, communication and improvement as a way to inspire you to articulate your purpose.
At this point, you may ask: why bother? Ideally, you just ask what your 'customer' wants and just start modeling.
Well, there's one key element everybody needs to remember. The purpose of modeling forms the foundation to which you would apply things like the scope of modeling, depth of modeling, the use of architecture and so on. In fact, the purpose of modeling is the defining factor to come up with our process modeling standards and conventions. These are important to produce consistent process models regardless of the modeler. They also form the consistency of how processes are visualized and consumed.
What happens without a modeling purpose?
Several things happen :
- There's unpredictability regarding the effort required in process modeling. And without effort, it's very difficult to identify when you finish and when you come to some form of conclusion so you can go onto the next phase.
- The second thing is that the granularity of process modeling becomes ambiguous. You could model all the way down to the detailed task level and not even know that that's the end.
- Simply put: without the purpose, there is no guidance.
Start your process modeling journeys with a strong and documented purpose.
Watch the video:
I believe you missed one very important reason to model: automation. And by that I do not mean as specification for someone to code, but in a model driven development environment. By modelling for that purpose, you also solve some of your problems: knowing when it is enough and also what the desired granularity is. if you have a tool that can read your models and instantly transform this into working software, the feedback loop to when you are done, whether or not it is modelled correctly, etc is much shorter. Also, there is less risk that your models won't be used, i.e. have no purpose.
Excellent. I would add : Target audience: know for whom you are modeling. JEM : just enough modeling, know thy scope and required details and depth Jim : just in time modeling : avoid too early models, as reality changes.