When Simulation Makes Sense

I’ve been on a mission lately to promote a different way of thinking about simulation, talking about Simulation’s Place in BPM and Why We Should Reposition Simulation in BPM. However, unlike others, I’m not throwing out the baby with the bath water: the key to my argument is that simulation is extremely valuable, but maybe not as a process design tool in the scope of BPM, as is the traditional use case for simulation.

Process Design

This sort of begs the question: does business process simulation ever make sense as a process design tool. Of course it does, just maybe not in most cases of BPM implementations. Here’s a list of cases where business process simulation might be appropriate as a process design tool:

1. The process does not yet exist.
2. True business process reengineering where you are obliterating your existing processes and starting over.
3. Where you are not implementing a BPM/automation system.
4. Where the costs of automation are very high.

The above cases make sense when the cost of simulation is much less than the cost of automation. Consider cases (1) and (2) where you don’t even know what you’ll be automating (if in fact you will even be automating the resulting process). These are strategic cases, and a good candidate for using simulation where many what-if scenarios can be worked on with no actual implementation costs. In case (3), business process improvement initiatives will always incur a data acquisition cost for simulation analysis as automation will never provide the data to drive simulations. I would think that case (4) doesn’t even require comment.

Operational Decision Making

It has been suggested that maybe we are still in the early adopter stage of business process simulation. That, to me, is simply not true. Business process simulation has been around for ages, and even within business process management, simulation is considered a key feature of BPM suites (think Gartner’s magic quadrant). While part of the problem might be the actual simulation tools provided with BPM suites, one has to ask the obvious question: if business process simulation makes so much sense, why isn’t everybody using it? And why aren’t they using it all the time?

It seems to me that if there is a cost associated with simulation analysis (and there most certainly is), then reuse of the analysis is critical. Process design happens once, the process runs all the time. The cost of simulation makes sense when it is amortized across its continual use in production. Simulation models can be used to fill in the gaps in production reporting when a process is not, or is only partially, automated. That allows managers to make better decisions. When a process is automated, simulation can be used to augment production reporting by predicting near and mid term process outcomes based on the current state of the process. Better yet, use the predicted outcomes to suggest optimizations to management, such as resource scheduling or even changes to the process structure itself.

In conclusion, of course there is a place for simulation as a process design tool. I just think there is a much, much bigger opportunity for simulation as a decision support tool in the actual ongoing operation of business processes.