Try business rules and predictive analytics for better decision management
By Peter Schooff, Contributing Editor, ebizQ
EDITOR'S NOTE: In this Q & A, ebizQ Contributing Editor Peter Schooff interviews consultant James Taylor about business rules and predictive
analytics in decision management. Taylor, CEO of Decision Management Solutions, is author or co-author of several books, including "Decision Management Systems: A Practical Guide to Using Business Rules and Predictive Analytics” (IBM Press, 2011). This Q & A is excerpted from a longer podcast; it has been edited for length, clarity and editorial style.
ebizQ: Where are companies at with decision management and predictive analytics today?
Well, it really varies. In some industries, this whole idea that you should focus on decisions [and] use business rules and predictive analytics together is well-established. If you go talk to a credit-card issuer or someone else working in retail credit, then this is absolutely taken for granted; this is how you run your credit business.
In other industries, much less so. The idea of systematizing their decisions, knowing what they are, is all still very new. They’re adopting business rules for specific projects; they might be doing predictive analytics for specific projects. But this idea that you could take control of decision making within your organization—particularly these transactional decisions—is still a pretty new one.
And most of the folks looking at predictive analytics are still focused on what I sometimes call “predictive reporting.” They're trying to see how they can embed predictive analytics in their current BI infrastructure, rather than saying, “What are the new opportunities for predictive analytics in terms of automating and taking control of some of these decisions completely?”
ebizQ: What would you say is the key for effective decision management for a company?
The [new] book has four principles in it, but I think the key one of the four is to begin with the decision in mind. You can only really build decision management systems—and, I would argue, only really take full advantage of business rules and predictive analytics—if you're really clear what your decisions are.
In particular, [I mean] the sort of operational day-to-day decisions that you make in every transaction when you process a claim, or you process a loan, or you talk to a customer. You have these very high volume decisions that are made in enormous numbers, and yet they're not on most company's radar when it comes to thinking about decision-making. When they think, “I’ve to improve decision-making,” they're thinking about executive decisions and management decisions.
So the key for effective decision management systems is really to say, “These decisions that run the day-to-day business—they matter in the same way that our operational processes or operational data matter. We have to have control of them, we have to know where they are and how they work, and we have to be able to improve them over time.”
ebizQ: One of the big mantras for businesses today is customer service. How does this help a company’s customer service efforts?
A lot of the energy behind decision management systems at the moment is really in this customer-service, customer-treatment arena. If you think about it, every time you interact with a customer, you have to decide what you're going to say to them. You have to decide what offer you're going to make, you have to decide what data you're going to ask them for, you have to decide if now is a good time to try and get them to sign-up for your newsletter. Whatever it is, you’ve got these decisions that have to be made, and they have to be made in large numbers because you make them about every customer. They're made typically by people in your retail stores, people in your call center, by your website, by your automated e-mail system.
You have to take control of those decisions and use the decision management system so that everyone and everything that interacts with your customers is doing so in a way that reflects your company’s strategy about your customers.
ebizQ: How would you design a system, essentially, with a decision in mind?
What I find is that you have to start by, typically, looking at a business process or looking at a system and starting to walk through it and say, “Well, are we really interacting with somebody here? Are we really moving data? Are we really capturing new information? Are we really just deciding what we're going to do next?”
And [then you] gradually focus in on those decisions where you're making a choice, a selection from a range of options. So at this point in the process, I can do one of a bunch of things and I have to pick which one I'm going to do. Well, that's a decision point. And having gotten that high-level decision—which is often something fairly generic like, “What should I do about this claim?” or “What action should I take with this customer?”—you have to decompose that a little. It's one thing to say “I’ve got an order-to-cash process;” it's another thing entirely to decompose that into its steps.
It's a little bit like that with decision-making. It's not terribly hard to find those high-level decisions that you make in the middle of this process. But to understand them and to build a system for them, you have to decompose them and say, “I can't make that decision until I've made this other smaller decision.”
So gradually, you iterate through that process, understanding the decision-making elements that go into the decisions you’ve got in your process. Then you say, “These bits of the decision—those are the bits I want to build in the decision management system,” because often the high-level one is not something people are 100% comfortable automating.
Let me give you an example [involving a commercial insurance company]. If I say, “Why don't we automate how we underwrite this commercial insurance-policy decision,” most insurance companies are going to say no. They're going to say, “I want to show that to an underwriter.” But if I decompose that decision, then they're sub-decisions, like “Do we have all the information we need about this location? Does the location need an inspection? How risky is this kind of business in general?” Those smaller decisions are often much easier for someone to say, “Oh, yeah, the system could do that bit.”
So you have to break down these decisions to see which bits make sense to go into a decision management system for your organization, because every organization is different in its tolerance for automation in this area.
ebizQ: I'm hearing some similarities between decision management and process management. Can you elaborate on decision management works with process management in a company?
Sure. My view is the decisions and processes are peers. They're two distinct things that you should have control of. Just as you should have a model of your business processes, you should have a model of your business decisions. The reality is that at many points in those processes, you have to make a decision.
One of the challenges we have with business process is that the business- process environment we gave our business users was, in many ways, the first time they had a way to look at their business and understand it. So that's their hammer and now everything looks like a nail. [Meanwhile, decisions] have gotten wrapped into the business processes. They're making the business process more complex and they’re making it harder to change, because they're not the same as a business process.
And so if you can extract those decisions and say, “No, I’ve got this decision; I'm going to model that and manage that as a thing. And I've got this business process that needs the answer that that decision is going to give it; it's going to need that question answered that's represented by that decision.” Then I simply find my process. It's much more streamlined. I'm much more likely to get sort of straight-through processing because, when I get to a certain point, I can simply call the decision-management system and say, “Can you answer this question for me,” get an answer back and keep going.
So by considering them as peers, you can simplify existing processes and make it a little easier to build fairly linear, straight-through processes.
[Often,] you talk to someone they say they have a standard process, but they have a couple hundred local exceptions. Almost always, those local exceptions are not really to the process, they're to decision-making. So if you separate out the decision and encapsulate the rules about these local variations inside a decision, then the process becomes, actually, genuinely, a single global process.
So there are a lot of synergies between the two, but you have to do think of decisions and processes as peers working together collaboratively to run your business.
ebizQ: What do you see ahead for decision management?
Well, I think the business rules community has very much decided this is a better way of talking about business rules at an automation level. I think you're going to see it become the norm, essentially, in business-rule [and] technology circles, to talk about decisions and managing decisions.
And I think you'll start to see some of the repository tool and the simulation tools and so on that are in those products reflect the more decision-centric mindset.
But I think the big interest is going to be in analytics. There's a lot of energy around analytics at the moment. When we look at the stories about who's been successful with predictive analytics and many of them—almost all of them, I think—are successful because they can apply those predictive analytics to the way they run their businesses every day. They're not getting one big “aha” moment from their analytics. They're getting improvement every day in how they interact with customers, how they manage risk, how they manage fraud.
I think that realization is going to spread as more people get interested in predictive analytics. It’s going to make them realize that it does no good to be able to segment your customers, or predict which ones are going to turn or which ones might like which offer, if your systems don't allow you to differentiate how you treat them. You're going to have to build decision management systems in order to give yourself the ability to decide differently about every customer.
Does your company use predictive analytics or DCM? If so, ebizQ editors would like to hear about your experience. Contact Site Editor Anne Stuart at email@example.com.
About the Author
Peter Schooff is a former contributing editor for ebizQ, where he also managed the ebizQ Forum for several years. Previously, Peter managed the database operations for a major cigar company, served as writer/editor of an early Internet entertainment site and developed a computer accounting system for several retail stores. Peter can be reached at firstname.lastname@example.org.More by Peter Schooff
ebizQ is the insider’s guide to next-generation business process management. We offer a growing collection of independent editorial articles on BPM trends, issues, challenges and solutions, all targeted to business and IT BPM professionals.
We cover BPM standards, governance, technology and continuous process improvement, as well as process discovery, modeling, simulation and optimization, among many other areas. We follow case management, decision management, business rules management, operational intelligence, complex event processing and other related topics. We closely track important trends such as the rise of social BPM, mobile BPM and BPM in the cloud. We also explore BPM’s use in functional areas, such as supply chain and customer management, and in key verticals, such as financial services, health care, insurance and government.
ebizQ's other BPM-oriented content includes podcasts, webcasts, webinars, white papers, a variety of expert blogs, a lively online forum and much more.