Thinking Analytically

What do predictive analytics, real-time monitoring, in-memory processing, and SaaS have in common?  According Doug Henschen at InformationWeek, they’re all a part of next-generation business intelligence.

Next-generation BI has arrived, and three major factors are driving it: the spread of predictive analytics, more real-time performance monitoring, and much faster analysis, thanks to in-memory BI. A fourth factor, software as a service, promises to further alter the BI market by helping companies get these next-generation systems running more quickly.

Interesting article with a number of good points.  What’s your take?

See the original post here. The post describes an interview with a birder and a doctor. They discuss similarities between pattern recognition in bird identification, medical diagnosis and business. Here’s an insightful quote:

HBR:  Does randomness have a role to play in birding?

Sibley:

Even then these vagrant birds (ones not expected to be in a certain location) tend to show up in places that have a pattern of rare birds turning up.  So they’re probably not random occurrence but a pattern we haven’t worked out yet.

ThinkingAnalytically discussion starters:

• Can you think of a time that an apparent set of random data turned out to be an undiscovered pattern?  What was it?
• How do you avoid claiming a random situation too early?

Use the comments to contribute.

Purchase full the HBR article here.

This amazing, interactive visualization was put together by the folks at the NY Times. It diagrams communication patterns of Presidential debate participants through mid-December 2007. In a nutshell, a line is drawn from one candidate to another each time the first candidate mentions the name of the other during the debate. To fully appreciate it, checkout the original and play for a bit. There’s a tremendous amount of information in one graphic including:

• Number of debates a candidate participated in (represented by slice)
• Political affiliation (represented by color
• Number of references a candidate made to other candidates (number of outgoing arrows)
• Number of references made to a particular candidate from other candidates (incoming arrows)
• Timing of such references within a particular debate (location of out going arrow)
• Timing of references over time (slices are organized by sequence of debates)

The technique is relatively simple but yet powerful. Though it may look like the output of my childhood spirograph, it is more like an inter-relationship digraph and generates interesting observations such as:

• Democrats apparently talk about each other much more than the republicans
• Democrat candidates that dropped earlier tended to be referenced fewer times (no surprise)
• Every candidate mentioned Clinton at least once

What do you see?

Thanks to A Beautiful WWW for bringing this to my attention!

This month, McKinsey Quarterly published eight technology-enabled trends they feel will help shape businesses and the economy. Trend seven is near and dear to my heart – or should I say head? Putting more science into management falls under the heading of leveraging information in new ways. The authors explore some of analytical topics that are all the buzz in recent best sellers: “ideagoras”, customer segmentation, experimentation, prediction markets and recommendation engines.

Leaders should get out ahead of this trend to ensure that information makes organizations more rather than less effective. Information is often power; broadening access and increasing transparency will inevitably influence organizational politics and power structures. Environments that celebrate making choices on a factual basis must beware of analysis paralysis.

The close by suggesting the following books:

• Thomas H. Davenport and Jeanne G. Harris, Competing on Analytics: The New Science of Winning, Boston: Harvard Business School Press, 2007. (see my entry on wikiBookNotes.org)
• John Riedl and Joseph Konstan with Eric Vrooman, Word of Mouse: The Marketing Power of Collaborative Filtering, New York: Warner Books, 2002.
• Stefan H. Thomke, Experimentation Matters: Unlocking the Potential of New Technologies for Innovation, Boston: Harvard Business School Press, 2003.
• David Weinberger, Everything Is Miscellaneous: The Power of the New Digital Disorder, New York: Times Books, 2007.

I’ve only read the first.  I’m interested in any reviews of the others.

What could be better than paying lawyers less money for more work? (Sorry Pat.)

In today’s world, litigation typically involves the review of thousands, if not millions, of e-docs. Though shredders can permanently destroy a paper document, past e-mails and e-docs hide in all types of digital nooks and crannies. Unfortunately, many companies have not kept up with the proliferation of e-documents. But that’s no excuse when in the middle of litigation. If there’s a chance that a relevant e-doc exists, it typically must be found, reviewed and deemed relevant or not.

Sean McNee describes a recent case where a small legal team representing a small time inventor was able to use visual analytics to review and analyze the contents of 50 million pages of electronic documentation in just a few months. (Yes, 50 million pages.) He states,

“Visual analytics tools identify nouns and noun phrases in a series of messages, then visually cluster the documents together according to similarities in subject matter”

Apparently some Fortune 1000 companies have seen three-fold productivity increases using visual tools. According to the article, “U.S. corporations spend nearly $5 billion a year analyzing emails for litigation, regulatory requests and investigations.” A three-fold increase in productivity yields signficant benefits!

Read more at BI Review Online.

In the latest round of tech acquisitions and mergers, IBM snatches Cognos. According this BusinessWeek article, this is a combination that might actually work. I think all of these big player acquisitions are creating havoc in large companies that have tried to standardize on enterprise solution providers. For instance, my company is clearly an IBM shop. No Oracle spoken there, but we’re a big user of Siebel and PeopleSoft. Oops, those are now Oracle products. Does that mean that there will be advantages in the future for Oracle users? I’m certainly no expert in the arena of enterprise software but it appears to me that the end of best-of-breed certainly has some downsides.

Technorati Tags: business intelligence, acquisition, ibm, cognos, PeopleSoft, Siebel