Transitional BI to Give Way for Change

Bangalore: It has been said that the old BI is giving way for a new order of data analysis. A new report titled “BI is Dead! Long Live BI” by Constellation Research analyst Neil Raden states that BI has reached a phase of transformation where it has become only one link in the “continuum of decision-management capabilities."

Raden states that this continuum will include everything from predictive modeling, machine learning, and natural language processing, to business rules, data visualization, and what the report categorizes as "traditional" BI.

A lot has been talked about how the BI adoption rates are below 20% or even 10% of potential users at many enterprises. This is because there are many limitations to the BI tools that were available earlier. Doug Henschen writes on InformationWeek that there is a need for better ease of use, ease of deployment, affordability, and ease of administration.

What is notably conspicuous in BI according to Raden, is the option for business users to create their own data models. Raden lists data-modeling best practices that he states will lead the way to improved BI. There is also a need for visual modeling tools that reduce complexity. Raden also calls for robust collaboration and workflow capabilities which will reduce rigidity and allow insightful decision making.

The latest BI and analytics market share stats from IDC show that the 3 fastest-growing vendors in the industry are Tableau Software, QlikTech, and Tibco Spotfire, with growth rates purported to be 94%, 43%, and 23% in 2011, respectively. These companies provide data visualization, analytics, and high-scale in-memory analysis capabilities. Larger vendors are catching up and imitating their interfaces and approaches.

IDC predicts that advanced analytics (for predictive modeling and machine learning) will rise at 10.1% per year through 2016 and content analytics (the parent of natural language processing) expands at 14.5% per year through 2016.

IDC predicts that open source, non relational Big Data platforms such as Hadoop and NoSQL databases will mostly run alongside existing business analytics systems, but in a few cases will cut into conventional data warehousing sales.

The demands of big data analysis make flexible data modeling all the more crucial. Business professionals want software that lets them "address only the meaning of data--not its structure, location, or format," Raden explains.