Nine Trends: One Failed Economy
January 7, 2009
Here is rural Kentucky even my neighbors are economizing. Instead of shooting squirrels with semi-automatic rifles, the Hatfield-like brood has reverted to snares. Most of the lower quartile crowd with whom I spend my time did not see the economic meltdown coming. In reality, the whiz kids in London, Manhattan, and Tokyo didn’t seem too aware of the problem either. What’s wrong? There are real time intelligence systems available from big name outfits (Thomson Reuters), smaller tune (Connotate), and newcomers (FirstRain, Silobreaker), among others. These systems deliver business intelligence, often accompanied by graphs, heads up displays. The Cluuz.com system takes the woeful Yahoo search system and makes it yield useful information.
I read with interest the nine business intelligence trends that a reader forwarded to me early this morning. A capable analyst (David Stodder) tackles this subject in the economic climate of the moment. I wonder if the present economic climate is a consequence of business intelligence or a lack thereof. Never mind. That’s a topic for another forum.
Mr. Stodder’s article is a fairly long one. I don’t feel comfortable summarizing his nine trends. Please, read his full presentation here. I propose to highlight three of the trends he’s identified and offer a couple of comments, fresh from my small, stagnant pond in rural Kentucky.
One of his nine trends is that “users demand a richer experience.” His point in that interfaces have to be more like dashboards. Users want to get answers and “integrate at the glass”. I agree. If you take a look at the typical interface for Clarabridge or Cognos (to pick two vendors whose systems I examined at a recent conference), I needed a statistics Sherpa to help me get from A to B. I am not the brightest goose in the flock, but it was clear that these two vendors make some big assumptions about their users. This user thinks the vendors’ assumptions are wrong. No wonder there’s an economic meltdown. The outputs can be polymorphic.
Mr. Stodder creates a “6, 7, & 8” megatrend that seems okay to me. The bundle struck me as a bit confusing. I think the idea is that the idea of “breaking” the “mold” is spot on. The idea is that getting real time information is important. Getting data more quickly means more capital expense. He is right there. What I think is missing is explaining that the capital investments can be non linear, which to me is an important issue.
The third item I found interesting was the reference to the Google. Mr. Stodder indicates that the traditional database architecture may not be up to the task of today’s business intelligence task. Again, he is on the money. I think that the cost of making the old style databases handle today’s petascale data management needs contributes to decision meltdowns. It is tough for me to comment on likely outcomes when I can’t process the complete data stored in traditional database tables in the time available to me to produce an answer.
You must read his other points. You may, as I did, find them thought provoking.
In closing, let me make a couple of goose-grade points before I lose them in the pond scum in which I am floating:
- The phrase “business intelligence” like “enterprise search” and “content management system” are essentially meaningless. What these areas comprise are shopping carts into which one can toss various technologies with labels that assert “money saving” or “increased efficiency in operations” or other types of marketing ?????? (shibboleth). These are rallying points, not solutions.
- The notion of making an entitlement grad or trophy MBA smarter with a dashboard that presents what he or she needs to know is crazy. The heads up displays in fighter aircraft become very simple when the action gets hot. The notion of mashing, displaying, and combining is crazy because it is usually difficult even for an expert to point out what is important. If these systems worked, do you think the Madoff issues would create work for hundreds of attorneys? Nope, interfaces are secondary to delivering accurate, meaningful data. In my opinion, business intelligence systems have a long way to go. If these worked, would the surprise attack on Gaza caught the authorities by surprise?
- The delicate suggestion that Codd style databases don’t do the job soft peddles a very serious problem in data management. The notion that Google has a good data management in MapReduce is like describing New York City in terms of the deli at Lex and 33rd. Google has blown past today’s “business intelligence” vendors. I don’t think the slumbering business intelligence industry, its pundits, or its most vociferous supporters realize that Google is moving and may put the key players in a position of check in Google’s digital chess game.
I might be wrong, but I think some shakeups, business failures, and technology marginalization will make the 2009 business intelligence landscape quite interesting.
Stephen Arnold, January 7, 2009
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
