≡ Menu

Opening Keynotes at #iod11

Share

Jeff Jonas‘ great ad for analytics, talking about crossing a road as an analogy for the power of predictive analytics, kicked off the first real content of the keynote. Jeff talked about enterprise amnesia, the challenge of enterprises being unable to process all the data they have flowing in. Jeff uses a puzzle metaphor and argues that companies don’t know even which puzzles they have pieces for, whether they have duplicates or missing pieces. He had a wonderful story of a group of kids working with pieces from multiple puzzles and how they worked through it. He uses this to describe the importance of context and the fact that context accumulates gradually though not in a linear way. Big Data in this view helps drive this improves context and makes it easier to find problems and assemble/use the data you have. Enterprise Intelligence involves not only building this context but also deciding, sometimes in less than 200ms sometimes by reaching out to a decision-maker. But this data also drives longer term “deep reflection”. “Sense and Respond” and “Explore and Reflect”. You need streaming data handlers, analytic appliances to process lots of data for deep insights, BI tools for people and Decision Management Systems based on rules and embedded analytics.

Sarah Diamond came up next to talk about the financial services industry. “Transformation is the new normal” in this industry and this requires a new approach to analytics and transparency. SunTrust came up to follow up on this to talk about their journey to a new approach to risk analytics. They went from multiple redundant areas to standardized reporting, from Excel and Access to an integrated warehouse infrastructure, from paper reports to portals. IT support is up, inconsistencies are down and data is managed more effectively. A massively more efficient BI environment for sure that helped SunTrust keep better track of what was going on as the markets went into crisis. Personally I think what they did is PART of the solution for risk management as they improved their ability to MONITOR risk but not manage it as operational decisions were made. It’s a beginning, a foundation, not an end.

The efforts SunTrust have made to make analytics pervasive are impressive but they seem overly focused on presenting data to people using BI tools. Personally I don’t see that as a route to pervasive analytics – only embedding analytics in Decision Management Systems that anyone/everyone can use will make analytic decision-making truly pervasive. A teller is just not going to have the time or skills to use analytics the way the knowledge workers and risk managers in the organization do. SunTrust are trying to embed analytics in their workflow and processes but it still seems focused on presenting it to people not empowering the systems themselves to make good decisions. Their closing point though, that analytics must be embedded in operational processes to be successful, is the right lesson.

Some great customer stories for predictive analytics followed with strong ROI but too little detail on how these analytics embedded into the operational processes and systems that drive the business.

Leaders, Mike Rhodin of IBM said, are those who can embed analytics throughout their business. They need to do this despite exploding data volumes and new sources of data. Meanwhile change continues to be a challenge and the gap between those making the most of analytics and their competitors continues to grow. Companies need systems that are agile (to respond to change), analytic (to make the most of their data) and adaptive (to maximize value by learning over time) – they need Decision Management Systems! As Mike said, you need to deploy analytics into their business processes to optimize outcomes. I don’t think this is mostly about getting analytic tools into the hands of people throughout the organization so much as getting analytic decision-making to everyone (which is not quite the same). Too many examples of putting reports and dashboards into the hands of people and of using predictive analytics only to route transactions to the right people rather than actually having systems that ACT on their behalf. Taking control of your information is good anticipating and shaping outcomes is good too but the critical step too many people miss is turning this insight into ACTION.

Robert LeBlance talked about IBM’s view of a sufficiently “bold” information strategy. It mus allow you to take unstructured content and structured data and bring it all together with integrated, federated, real-time management and allow for Big Data which means that external sources are as important as internal ones. He did not talk about being bold enough to embed analytics in your systems which is a pity.

Arvind Krishna came up to talk about some new IBM announcements:

  • DB2 Analytics Accelerator (using Netezza) and IMS 12 for improved performance
  • Content and predictive analytics for Healthcare and IBM Case Manager 5.1 for medical insights and advanced case management
  • InfoSphere MDM – a unified MDM platform and new capabilities for integration and governance for big data sources
  • iPad support for Cognos and new personal analytics tool
  • Algorithmics now part of IBM for risk monitoring and analysis
  • Q1 Labs is going to be acquired for security management
  • InfoSphere BigInsights and Streams for handling streaming “big” data
  • New Business Analytics and Optimization Jumpstart services

Lots of good stories, lots of good advice but not enough on how all this actually hits the front-line systems and people – not enough on Decision Management Systems and operational decision making.

Share

Comments on this entry are closed.

  • Paul Vincent October 25, 2011, 6:02 pm

    “Big Data in this view helps drive this improves context and makes it easier to find problems and assemble/use the data you have. Enterprise Intelligence involves not only building this context but also deciding, sometimes in less than 200ms sometimes by reaching out to a decision-maker. But this data also drives longer term “deep reflection”. “Sense and Respond” and “Explore and Reflect”. You need streaming data handlers, analytic appliances to process lots of data for deep insights, BI tools for people and Decision Management Systems based on rules and embedded analytics.”

    +1 – interesting to see the relation between CEP / sense and respond and long-term analytics (big data) and of course both feed into operational decision management…