IBM made a big announcement about Decision Management today. They are marking what they see as a new era of computing, cognitive computing that allows us to build systems that learn and, in my words, make decisions. All these new technologies, analytics and more, are about making better decisions. As Mike Rodin put it, the explosion of new data sources and increasing data volumes create both challenges and opportunities. A new decision management platform enables companies to put this data to work improving decisions at every level. One example of this platform was an insurance company that got a 70x improvement in claims processing while also eliminating a major fraud problem.
So, Deepak Advani came up to introduce the new platform. Recent studies from IBM show that 75% of CIOs with mandates to transform are looking to drive better decisions while 84% of outperforming CEOs strongly agree that translating insight into actions helps them differentiate their business. Like me IBM sees decisions from strategic and one-off decisions through monthly, weekly, daily and increasingly operational decisions. These decisions need to be made in real-time and at the point of contact with the customer. Not just about helping knowledge workers make better decisions, about making your systems and devices and front-line staff make better decisions. Lots of examples across every industry from insurance to supply chain to marketing and more.
Decision Management and the associated technologies have been used to automate decisions for many years. Increasingly the combination of these techniques can be used to improve decision making. For instance
- Business rules have been used (by Visa for instance) to process credit card settlements by applying 30,000 rules that change daily. And Predictive Analytics could be embedded to detect fraud as transactions are processed
- Predictive analytics has been used (by First Tennessee for example) to cross-sell and up-sell products. Optimization can be added to help target limited budgets and manage trade-offs between the various offers.
- Optimization has been used (by Continental Airlines) to improve crew scheduling. Business rules can be added to manage country by country regulations on crew management.
This is what IBM sees as the driver for the new platform – a need to combine these different approaches.
Why now? Well IBM sees the emergence of big data, a shift of power to consumers and the ongoing pressure to do more with less driving an increase in need for Decision Management. Doing more with less means automating more, making systems able to manage more of your business. Customer-centricity means making better decisions about how to treat customers. And Big Data is only useful if it can be turned into better decisions, especially the kind of repeatable decisions that generate all this data.
With that Deepak moved to some specific examples:
- Santam, an insurance company, using business rules and predictive analytics to identify claims fraud. 70% of claims are now settled within 48 hours while and this efficiency plus more fraud detection means Santam can reduce its rates while delivering better customer service.
- C spire wireless, a wireless service provider, is using predictive analytics and business rules to drive personalization – a differentiated, personalized wireless experience across all channels.
- Preventative maintenance is an example of using all three techniques – business rules about known issues and repair sequences, predictive analytics to predict failures and optimization for managing repair and maintenance schedules
So, IBM sees successful Decision Management requiring the combination of business rules, predictive analytics and optimization. And IBM likes to be a reference client for itself so one of the groups using Decision Management is IBM’s outsourcing group where it uses it for mortgage decisioning.
The platform builds on event and data inputs to a core Information Management Foundation to manage information. The new IBM Software for Decision Management delivers Decision Services that can be embedded into business processes, solutions, applications, existing systems etc. Then of course you need a feedback loop and governance process to manage continuous improvement. The platform contains:
- Business rules, Business Events (IBM WebSphere ILOG Operational Decision Management)
- Optimization (IBM CPLEX)
- Predictive Analytics and Text Analytics (IBM SPSS)
- Entity Analytics and Sentiment Analytics (new IBM SPSS features)
IBM sees this as a new category that creates the potential for pushing these capabilities into every industry solution, packaged applications like Unica, combined with streaming data management and in-line scoring/decisions for high performance real-time (IBM InfoSphere Streams) and pushed onto z/OS to support IBM’s historical processing systems.
Decision Management from IBM then is integrating all these different approaches to decisions, innovating with new capabilities and integrating it everywhere. And then there’s Watson….