|
Applications: The Democratization of BI
As much as any technology out there, business intelligence (BI) software has crossed into uncharted territory: in-memory analytics, visualization, integrated search. "We are getting to the point where the historic divide between business and IT will collapse," explains Andrew Bartels of Forrester Research Inc. in Cambridge, Mass.
BI has traditionally been defined as a set of IT tools that give business people insight into their organizations, typically with a lag. There is value in a rear-view mirror. A remembrance of things past can inform on the present, certainly, if the data is good. Our prediction: The Proustian approach of analyzing large sets of data after the fact is giving way to highly specific, special-purpose tools that can look ahead and guide activity in a proactive fashion.
To do so, the intelligence must be embedded in the applications, not sitting outside and simply capturing emerging information. Embedded intelligence requires a deep understanding of cause and effect specific to an industry, in a specific business context. "You have to be able to take that knowledge and embed it in a set of rules that say, 'If this happens, then do this,'" Bartels says.
Vendors haven't missed the signs. Santa Monica, Calif.-based Accruent Inc., a software provider that uses an embedded BI engine from Business Objects SA, shifted its focus from retail software to software specifically designed to help retailers get the most out of the real estate they own or lease. Model N Inc. in Redwood Shores, Calif., provides analytics that help the pharmaceutical industry and others calculate the mix of discounts and rebates that will generate the most revenue.
Analyst Kurt Schlegel, who writes on BI for Gartner, agrees that the view of BI as an IT-centric activity -- getting the right information to the right people at the right time -- is outdated. "BI has to evolve beyond that and move toward business processes and business strategy," Schlegel said.
Traditional BI required IT to jump though a lot of hoops, collecting and cleaning up data and putting it in reliable formats that could be queried against. An emerging technology such as in-memory analytics says "to hell with that," Schlegel says. With in-memory analytics, "you don't summarize, don't pre-aggregate, just take all your data, compress it, load it into memory and make your queries." The declining cost of memory and low-cost 64-bit computing enable in-memory analytics to surpass typical disk-based BI deployments. Indeed, midmarket companies are jumping on the in-memory analytics bandwagon.
Vendors such as Applix, QlikTech International AB and Spotfire are poised to compete in the broad BI platform market, Schlegel says. "A lot of the emerging technologies are attractive for a reason; they have good usability and break the IT bottleneck." However, Schlegel tempers his comments with a caveat: "If done incorrectly, you have no control and no standardization, and you're in trouble."
|
 |
|
|
