When enterprise search vendors say they're moving into Business Intelligence, I walk to my closet to take out my skepticism hat. Wearing that, I tend to think this is usually marketing hot air, or, at least, stretching the definition of what BI actually is.

So I'm happy to see Exalead's Paul Doscher explain this as database-offloading: rather than have the database perform complex queries it's not particularly good at, you can use a search engine (and its own index) to take care of them. He gives the example of a track-and-trace application used by a transport and logistics firm, where the search engine is much quicker at providing the information than the database itself would be. (It can also be a great cost saver. I've spoken to several banks and insurance companies that rent processor time on IBM mainframes, and they actually save money deferring as many queries they can to a search engine they run themselves, using a separate index of the same data).

That's a useful way to employ a search engine. So it's a bit of a shame that while this clarifies what "BI" means to Exalead, the same blog post rather needlessly drags in references to the MapReduce and Hadoop hype (we cover Hadoop in our Lucene review), linking this to Exalead's DSort. While in some scenarios, this might be useful technology, there is no direct link between BI and MapReduce and I don't think it's particularly helpful to customers to explain one buzzword and immediately introduce another.

Of course, I was kidding about walking to the closet to take out my skepticism hat -- in reality, I wear it all the time, particularly when dissecting hyperbole.