Is It the Right Time for Real-Time BI?

Ann All

What good is business intelligence if it doesn't enable companies to make decisions at the elusive "right time?"


Not much good at all, judging by BI vendors' efforts to speed the query times of their products. The direction is clear from the products' names: Teradata's Active Data Warehousing and IBM's Dynamic Data Warehousing, for example.


A key, according to this column, is substituting in-memory analysis for the traditional online analytical processing (OLAP) approach to queries. The columnist calls in-memory analysis "one of the most innovative areas of BI today."


Of course, she adds, not all companies are ready to capitalize on real-time BI. Problems with data quality and with process -- ensuring workers know how to act upon real-time insights -- will need to be solved first, she says.


Service-oriented architecture will help pave the way to real-time BI, say experts quoted in this Insurance & Technology Online article. A distributed, Web-based environment could ultimately replace the centralized data warehouse, some say. This change, assuming it comes, is many years down the road, however.


In the meantime, companies will continue to spend big on existing BI products. North American firms will lay out more than $23 billion this year, estimates AMR Research, a 9 percent increase over 2006. The market is "nowhere near" a saturation point, says an AMR Research analyst.

Add Comment      Leave a comment on this blog post
Apr 4, 2007 1:19 AM Joseph Pusztai Joseph Pusztai  says:
Applix TM1 has been an in-memory MOLAP engine since its inception in the late 1980's. It is one of the fastest OLAP engines on the market.I'm not sure what the author means by a "traditional approach to OLAP" - if it means disk-based ROLAP, then I agree that is usually very slow and clunky and requires "bolt-on" accelerators. Reply
Apr 5, 2007 8:22 AM Joseph A. Montione Joseph A. Montione  says:
OLAP Cube Processing; is the data stale as soon as you push the create cube button?The author makes a great point by focusing on the problem with OLAP'S predominant implementation architecture; when referencing, traditional approach to OLAP, in my opinion.Being that most OLAP architectures create cubes of data at a point in time, I find the authors comments interesting and insightful.We will find this to be an area of much advancement in the coming years. The vision here is to have the cube process itself incrementally as the source data changes.Why should BI professionals have to wait until the data is cooked into the casserole of what is todays data delivery flavor of the week?Joseph A. MontioneFounder: GTOpendatabase L.L.C.A Solutions Provider Reply

Post a comment





(Maximum characters: 1200). You have 1200 characters left.




Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.