Just read Sandy Kemsley’s comments about Rashid Khan’s blog about simulation & optimzation hype.
While a lot of BPM simulation tools are difficult to learn how to use they’re still incredibly valuable and can provide valuable information including:
- Base line time & costs associated with the ‘as-is’ state of a process
- Rational way to evaluate multiple possible ’to-be’ scenarios
- Data necessary to create a project ROI & payback
- Data necessary to compare actual production data vs expected results in from the process model. Round tripping.
- Presenting the business case to users, stakeholders and management
I don’t think there’s any need for us (BPM vendors) to hype modeling and simulation capabilities. A good analyst that understands the limitations of modeling and simulation can still provide incredible insight into process performance that’s not possible without these tools.
As far as the ‘task time’ issue. It is possible to come up with the expected time for a person to complete a task. For example, a part time administrator that works from 10:00 to 2:00 (with a 30 min lunch) should still only take 10 minutes to fill out a few fields on a form even if they stop to take some phone calls, drink their coffee, etc). So if a task was sent to them at 3:00 the previous day, and sat in their inbox until they opened their task at 10:01. The ‘step time’ may be 19 hrs & 11 minutes, it still only took them ~10 minutes to complete the task. You do need a way to differentiate between duration of a step vs the actual time it took to complete just the task. In a model you can define these values. To identify the actual task time, you can build clocks/ date stamps,etc. for fields on forms to track time or prompt a user to provide their estimate of what it took them to complete their work. Yes they might lie.
Nobody said that its easy. It’s NOT.
The real question is will the investment in time and development needed to capture the simulation results worth the effort? What’s the least amout of data you need to capture to be able to validate your assumptions & results?
