I have often posted on the need to combine decision management and process management but it seemed to me that recently I have seen more BPM writers talking about this. For instance the folks over on the ARIS blog posted BPM + BRM = Greater than the Sum of the Parts (talking about a webinar from Sandy Kemsley) and Tom Baeyens presented on integrating jBPM and JBoss Drools prompting Paul Browne to post on How to combine Workflow and Business Rules – in 5 easy steps. Meanwhile While Gary posted The kittens and the business rules.. which was definitely the funniest. Will all these posts out there I thought I would write a longish one on the topic.
One of the problems I see when organizations work on process management is that they focus on streamlining workflow and reducing handoffs. The resulting processes, though, remain dependent on human interactions to move transactions along. The primary reason for this is decision making – when a decision is required the process waits for a person to make the decision. Decision making also plays a hidden role, with decisions that are not recognized as such and are hidden in processes. Attempts to use BPM tools to automate these decisions can result in overly complex processes that are difficult to change.
Identifying and automating the decisions within a process will help process management move to the next level. Why? Because it:
- Eliminates manual reviews
- Allows decision making approaches to change independent of process changes.
- Puts business users in control
- Provides an ideal place within a process to add analytic insight
Organizations that adopt process management approaches and technology without considering decision management run several risks:
- They may automate an inefficient process based on manual review
- They may fail to build a process that can respond fast enough because it must wait for manual intervention
- They may be unable to change the decision making approach without also changing the process
- Long running processes may be tied to decision making rules that have subsequently changed
- The decision making logic is buried in the new process, making it hard to find and manage
- Personalizing or micro-targeting a process can be impossible because there is no decision point at which to personalize.
- Agility will be lower than expected because decisions cannot be changed independently of process
Organizations attempting to automate their business processes and achieve both straight-through processing and true business agility will need to do more. Decisions previously taken by people, or not made at all, must be replaced by embedded, automated decisions. Decisions previously hidden in process definitions or legacy code must be exposed and managed thoughtfully. Processes that used to wait for a person to be available and then rely on that person’s judgment must be replaced with fully automated processes that take advantage of compliant, accurate, automated decisions. To get the most out of process management, you need decision management.
Other links by me on this topic include:
Comments on this entry are closed.
Good post – to add (or repeat!) what you’re saying, there are two main areas where intoducing rules can help your workflow
– to automate manual decisions, in a key that those business decision makers can still understand what is going on.
– to replace existing automation – moving it from Spaghetti code to a set of rules that can be easily understood by business people.
Both add value, and both often have the advantage of capturing and documenting previously informal processes.
Paul
Hi James, I couldn’t add a trackback so I just linked my answer to your post (which was initially a small comment but grew into a real post). Please checkit on: http://marwaninaltum.blogspot.com/2008/07/reaction-on-edmbpm.html
Marwane
Now to another comment which is not directly related to your post. About EDM, BRM and BPM, my personal point of view is that BRM and BPI (process intelligence) techniques such as process mining, performance management etc… are the link between EDM and BPM. The first are the techniques and the latter is the discipline which is needed to be used in order to achieve efficient governance of the process-driven enterprise. I guess we should really go further now that still motivating the need for well-though methodologies for efficient use of BRM and BPM together, and actually profit from experience and come up with concrete methodologies. Of course, this is easier said than done, but I am convinced that the community should now move to the next step. You have a nice feature here that I also use for my blog which is the wiki, and I would propose we gather the points you made in the post with those made in the post by Kai-Uwe on the ARIS blog and use it as the start of a kind of a micro-knowledge base on BPM+BRM. What do you think? 😀
Regards, Marwane