All this week people attending one of the five InfoX conferences in New York City have been asking me if I’m blogging the conferences. The short answer is “no.” I couldn’t get high speed Internet access in my room and the Hilton was charging $2.95 for 15 minutes in the conference rooms. Plus, I was really busy with the conference itself. The first day I moderated all the sessions in the Business Intelligence Conference . Even if I’d been willing to spend the money for 15 minutes here and there, it seems rude to be sitting in front of a roomful of people typing away on a computer. In the audience is one thing, hosting the conference is quite different.

Here’s a brief recap of the first day of BizInt: I was trying for a blend of talks on business intelligence and competitive intelligence. The conference hoped to find some synchronicities between the two and to clarify definitions of the terms. KMWorld publisher Andy Moore’s keynote talked about how we’ve moved from some of his early concerns in publishing for the content world, which included considerations of stapler removals and the footprint of file cabinets. Yes, that actually made sense when he said it. He also told us we should think more about BI’s value proposition and less about ROI. That turned out to be a theme during other people’s presentations, although there was no consensus. Some people stressed value, others ROI, which I found very interesting.

He was followed by Knowledge inForm’s Cynthia Cheng Correia, who gave some very practical advice about choosing CI software. I loved it when she talked about “crack searchers,” by which she meant really good ones, not ones looking for drugs, and intimated that ONLINE would be their first choice of reading material. Thanks, Cynthia!

The afternoon saw joint presentations from Andrew Bernstein (CEO, Cymfony) and Edward O’Meara (CEO, MediaHound), talking about how to value the BI function. Reputation is all about conversations, but you have to listen to them, said Bernstein. O’Meara gave a case study on uncovering and analyzing B2B marketing intelligence, noting it’s a complex, lengthy process. I liked his comment about smaller companies’ struggle with data collection and analysis. Their answer: Make it up.

Then we went into BI analytics with Michael Schroek from IBM and Ian Scott from Angoss Software. From the perspectives of their companies, they discussed predictive analytics, how data, if properly analyzed, can provide some roadmaps to guide business strategy. Both ended up talking about the insurance industry’s use of predictive analysis to reduce risk, but agreed that some things, such as government actions, cannot be predicted because there’s no real data to analyze before the fact.

David Carpe, principal of Clew LLC took the audience back to the competitive intelligence viewpoint, talking about non-technological information/intelligence gathering techniques, particularly interviewing knowledgeable individuals, and gave some hints about where to look for information outside the U.S.

Finishing up the day was Manya Mayes from SAS and David Bean from Attensity discussing data and text mining. Manya gave some anecdotes about misunderstanding data, some from her present job, some from prior ones. She mentioned the marketing person at 7-Eleven who was so proud of her campaign that raised sales of Slurpies, until someone pointed out that Slurpie sale always go up in the summer. Mayes also had a lovely slide illustrating the dangers of textual misunderstandings when trying to effect text mining. Some of her points were: UPPERCASE, Miss-spelings, A.C.R.O.N.Y.M.S, Shrt-hnd, Prof@nity, and *Punctuation*. Brilliant! Bean subtitled his, “Elephants, the Electrical Powergrid and your 8th grade English Teacher.” His major point was that you can extract facts from unstructured text if you can “diagram” sentences as you were taught in the 8th grade and transform text to rows and columns. Some of his examples were quite striking, including analysis of government data on automobile failures. I don’t think I’ll be buying a Ford anytime soon.

The second day, ably moderated by my colleague at CRM magazine David Myron, started with a topic dear to my heart: Quality. Knowledge Integrity’s Richard Ordowich made a great case for the importance of high quality data. He echoed Mayes’ comments on what can be wrong with the data you’re using. He also talked about the importance of information compliance and how critical compliance and quality are to intelligence success. Another SAS speaker, Anne Milley, distinguished between insight and hindsight. Through a series of case studies, she showed that getting an answer is not enough, you need to operationalize analytical models and use analytics in multiple functional areas.

Anna Kotsalo-Mustonen came all the way from Fountain Park Ltd. in Helsinki, Finland, to convince us that we should learn to see beyond our comfort zone, and “brain mine,” not just text mine. We must understand the mental model of customers and partners for proactive decision-making.

Mike Brooks, Theoris Software and Temi Grafstein BetaWatch, Inc. covered merging internal and external data, while Richard Brath, Oculus Info showed some live demos of using visualization for business intelligence. Mark Anderson, TechSource Group, Inc., gave the final formal presentation on managing aging data. He suggested we not keep everything, much as we’re tempted.

Throughout the two days, there were also product presentations and demos from the six sponsoring companies of the BizInt conference: Autonomy, ClearForest , CoEmergence, Inc. , Fast Search & Transfer, Hoover’s Inc. , and ZoomInfo .

We ended the Business Intelligence conference with a panel discussion among Dave Myron, Andy Moore, and me, and took questions from the audience. If anyone wants my concluding slides, let me know and I’m e-mail them to you. Yes, I’ll include the picture of the Red Swingline Stapler and the detailed diagram of a staple remover, not to mention the stapler haiku I put up on the screen. All in all, a very valuable two days, which would have been better with more available high speed Internet access.