Rules, analytics, decisions!
Rules, analytics, decisions!
When Ron and Keri asked me to start writing a column, one of the hardest things to do was pick a title. In the end I was inspired by the movie business with its lights, camera, action! As anyone who has read my articles before, reads my blog, or has checked out the book Neil Raden and I wrote (Smart (Enough) Systems) will know I like to write about decisions. Decisions lead, of course, to actions — indeed this focus on actions is a critical aspect of the enterprise decision management or EDM approach. Where lights and cameras may be the key elements that must be readied in the movies, rules and analytics must be brought together for decision making. Quite coincidentally, the acronym RAD also stands for Rapid Application Development, something that EDM directly empowers.
So — Rules, analytics, decisions! it is. Well, almost. For one thing, there are many other writers on the BRCommunity discussing business rules so I doubt I will have that much to add — I am likely to simply try and put rules into the context of EDM. For another thing, the title omits 'adaptive control', one of the key concepts of EDM. I will be writing therefore mostly about analytics, adaptive control, and decisions but I like the title so it stays anyway.
BRCommunity recently finished running an extract of the book I published last year with Neil Raden — Smart (Enough) Systems. The 10 part series laid out the need for, and ROI of, what we called smart (enough) systems. Enterprise Decision Management is the approach that can deliver these kinds of systems most effectively. This column is thus an extension of the approach we documented in the book and an ongoing follow-up to the 10 articles.
One thing I should mention at this point is that the name, Enterprise Decision Management, has some competition. There are those that call it "Business Decision Management" and those that simply call it "Decision Management." Some talk about "dynamic business applications" or "intelligent process automation" and include the concepts of EDM within those broader terms. Others talk about becoming decision centric or making data-driven decisions or using automated decision systems. No matter, the concept is consistent no matter what it is called.
Here's the definition that Neil and I use when we talk to people:
Enterprise Decision Management is an approach for automating and improving high-volume operational decisions. Focusing on operational decisions, it develops decision services using business rules to automate those decisions, adds analytic insight to these services using predictive analytics, and allows for the ongoing improvements of decision-making through adaptive control and optimization.
Several things are important about this definition:
- It's an approach, not just a set of technology.
- It's about automating and improving decisions, not just one-time automation.
- It's about high-volume decisions, the ones that go on day in and day out and drive your transactional systems, not just the ones that management takes.
- It results in decision services — self contained components that make decisions.
- It relies on business rules and business rules management systems.
- It uses data mining, predictive analytics, and optimization to find the right rules and to make those rules "smarter" by inferring interesting things from the data you have.
- Ongoing improvement requires adaptive control, sometimes known as randomized testing or champion challenger testing.
If you are already using business rules and the business rules approach, you probably know what a decision is. You are almost certainly using business rules to describe how a decision should be taken. It may not be the only thing you can do with rules; in fact I know it isn't, but it is the thing you can do with business rules that has the biggest bang for the buck. If you are not using business rules, you might find decisions in your use cases or processes — it would be astonishing if you did not — but it might be less obvious to you. Focusing in on decisions, especially the decisions inside transactions and operational processes, is the first change that EDM requires and will be the topic of the next column I write.
If you are already focused on decisions and already using business rules, then the concepts of data mining, integrating rules with predictive analytics, and using adaptive control to conduct randomized tests will likely be where you want to see some focus. These will come up in subsequent columns, unless I get distracted by something interesting.
If you already know all this, then you are probably already using EDM and you'll have to wait a while before the column will get to things you don't already know. Alternatively you could just subscribe to the Smart (enough) Systems blog.
See you all next month.
# # #