BPM Strategies - March 2008 - (Page 18) B P M I N D U ST RY L E A D E R Q & A Taylor: They also have a very high rate of change as well. The triggers for change for decisions are not the same as the triggers of change for other parts of a business process. the “champion” decision from then on. So many things that make a decision good or bad can’t be told at the moment you make a decision and so many are impacted by behavior that is outside of your control what your customers do and what the market does - you need to have a much more aggressive approach to testing and refining decision-making. response system. If you are a customer it can pick up your phone number and know who you are. But does it change the menu options it gives you? No. Nobody thinks about that as a decision. But every time you offer someone a menu, make somebody an offer or send them a letter, you are making a decision about how to interact with them. The customer experience is the cumulative effect of all these interactions. But people don’t see all these decisions. There are very few companies who will try to predict what is best for you and then interact with you and let you say, “Yes, that is good. Let’s do it that way again,” and build a combination of their rules, your preferences and the analytics that reflect how people will probably respond to come up with the best decision or offer. BPMInstitute.org: How so? Taylor: The pricing my competitor offers probably does not change my process but it will change some of the rules and some of the decisions, because I want to be competitive. I might change my rules every day based on what other people are doing but my processes don’t change. These decisioning pieces have very much their own agenda. BPMInstitute.org: Can you summarize the approach? Taylor: The five elements of business decision management are focusing on decisions; building decision services for your platform, using rules and analytics to build those decision services and then using adaptive control to constantly improve them. BPMInstitute.org: What is the other conceptual change? Taylor: The most unusual aspect outside of the analytics is the idea of adaptive control. One thing that makes decisions unusual is that to know if a decision was a good one or a bad one often takes a finite and non-trivial amount of time. When I make a decision, sometimes I can tell if it is good or bad immediately. If I want to cross-sell you something and I do, I can tell if it was a profitable decision. But if I have five different offers and some are subscription-based, I may not know if the decision I made was in my long-term interest until long afterward. If I offer you a special two-year subscription instead of a one-year with renewal, I won’t know if that was a good decision until I see what happens with the people who got the one-year offer. The results of decisions can take a long time to play out. BPMInstitute.org: Everybody knows they should make good decisions. Why is it a challenge to convince companies to try to make good decisions? Taylor: There are a number of things. When I talk to IT folks, I find that business rules are well-known but not widely used. There is a perception that they are slow; that they don’t add that much value; that they are only good for writing expert systems code. Data mining and analytics are almost unknown among IT folks. So, the first problem is the lack of awareness of some of the technology and mis-perceptions about it. The second problem is the need to understand decisions. For instance, consider a company that is going to send a letter to get customers to come into its stores. If you ask the marketing person how many decisions they had to make, they would say four or five - the list, the pitch, the production values, the offer and so on. But if you send the offer to 100,000 people, I would say that you made 100,000 decisions. Each person will receive that letter and will react as if it were a deliberate decision on your part. It may be worth thinking about it that way so you can tailor it to each customer. So this is another problem - the decisions that will repay this kind of investment are often buried in applications in which every transaction is treated the same way. Another classic example is an interactive voice BPMInstitute.org: It seems hard. You need technological change and cultural change. How do you overcome the fear factor? Taylor: Some of the fear factor has gone away with the move to a services-oriented architecture. The pressure to become more event-driven and real time is making people do it. Since they can’t make a unique process for each customer, they have to make it feel unique to each customer. It is almost like a second order problem. After you have your business processes fixed; you have your data store; you are able to handle your transactions; and you have eliminated the mindless repetition, it becomes clear that this is the next option. We help customers find the decisions in their business process and determine who makes it, what data it is based on and what is the payoff for making better decisions. BPMInstitute.org: So? Taylor: You don’t know how well the approach you picked worked at the time you picked it. You have to take a subset of your transactions and apply different approaches. With adaptive control, you say that you are going to run 95 percent of the transactions one way and five percent with alternatives that may be better - say, one percent for each alternative. Over time, you can see which ones work and which don’t and one might even be better and become 18 BPMInstitute.org: What is the payoff? Taylor: The ROI is really very high. As more companies get more control of their processes they are starting to see the value. People initially get a huge ROI from their investment in BPM because their processes are such a mess and they clean them up. There is another ROI in taking a look at their applications and their decisions. How can I make more profitable decisions that are better for my customers? BPMSTRATEGIES http://BPMInstitute.org http://BPMInstitute.org http://BPMInstitute.org http://BPMInstitute.org http://BPMInstitute.org http://BPMInstitute.org http://BPMInstitute.org
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.