Jim Sinur took the stage for Are Rules Inside-Out in BPM?, where he claimed that he’d push the envelope in how we thought about rules. He started with how rules are a start, but agility requires a full business rule management strategy so that you can manage the rules that you’ve externalized, especially if you have multiple business rule engines. Now to be fair, many organizations haven’t even externalized their rules yet from their enterprise applications and business processes, but if they ever do, they’d better be ready to manage them or they’ll have a big mess of rules to deal with.
Today’s business rules landscape is pretty confusing, covering everything from neural nets and expert systems to business rule engines and business rule management systems. If these business rules are too rigid (unchangeable), it impacts the agility of the business processes and the entire organization; if IT has to spend a huge amount of time and money to change rules, then you can be sure that it’s not going to happen very often. However, IT is often unwilling to put control of the business rules into the hands of the business; there needs to be a way to have proper governance over changing of rules, but not so much control that it’s impossible to keep up with shifting business requirements. In many cases, the business has no idea how difficult it is to change any given rule, and some standardization of this — via rule externalization and management — would also improve service levels between business and IT.
The key is to understand where rules affect processes, and see where the ability to change rules for in-flight processes can greatly improve agility. Sinur went through the business benefits of rules, and some of the risks of fixed rules: primarily, business rule management is an enabler for governance. He also walked through different models for adopting rules: the safe and steady control model (slightly smarter process), the cautiously dynamic model (process with above average intelligence), and the aggressively predictive model (Mensa process). Obviously, the model has to suit the organization’s risk tolerance as well as the underlying process, since these range from just automating some well-understood decisions to suggesting and implementing new rules based on behaviour within the process.
He has some great recommendations for getting your rules under control, including such thoughts as 15% of the rules are the ones that the business really needs to change to remain agile, so pick the right ones to externalize, and understand both the business benefits and risks associated with changing that rule.
Watch for Gartner’s definition of what should be in a BRMS later in 2007, since this is becoming somewhat of a commodity.
One of the interesting things about bringing rules out of processes is the ability to embed more intelligence/analytics into the decisions – check out this post from a previous Gartner event on analytic process controlling for instance.
Many of us in the rules space have a pretty good idea of the difference between a rules engine and a rules management system and on what you need in a rules management system.
have fun at the conference, sorry I am not there (in Lisbon for Fair Isaac’s European conference).
JT
James, great posts — I think that most (likely including me) don’t really grok the distinction between a rules engine and a rules management system, but your post helps to clear it up.
The combination of business rules and business intelligence is killer in terms of improving decision, that much is definitely clear.