BI Product Landscape

   Business intelligence is not of necessity about tools and technologies; somewhat it is a plan of combining data from various sources with methodologies that make those facts harden in a unified manner.

   The data part of this strategy is data warehousing (see below Figure ). Once the data is sourced, scrubbed, enriched, conformed, and in conclusion housed in “access-ready” formats BI tools can make the data sing and dance.

   In above  Figure (adapted from Inmon, Imhoff and Sousa, 2002), BI is traditionally found in the Data release area (to the right side of the Information Factory.) Operational BI, on the other hand, is coverage and analysis done directly off of the operational systems or via an Operational Data Store (for example see, Imhoff, 2001.)

   Traditional BI makes use of past data points (what you know about the data from a historical viewpoint) and displays it for the end user to make main inferences. The historical reporting takes advantage of the dimensionality in the data to “slice and dice” by reporting facts along any number of dimensions.

   Early reporting tools allowable programmers to define exactly what they wanted to present in unreliable levels of granularity and aggregation. In the 1980’s a plethora of OLAP style data structures emerged, which included MOLAP, ROLAP and Hybrid-ROLAP. All of which provided the ability to drill in, around and through to make sense of the data presented.

   While OLAP is certainly not dead, highly structured interfaces to the data came out of an organization’s executive branch interested in the details. In other words, taking data from “green bar” and simply transferring it to the “browser” was not sufficient.

   Management required to manufacture the data into meaningful bits of information. “Tell me what’s wrong. Highlight the facts for me,” was the driving force at the back the dashboard and scorecards in today’s electronic toolbox.

   coverage on the past can only show what has happened, not what the future may bring. Past information must be joint with some real-time information and then covered with analytics in order to have true foreknowledge. This is where data mining, forecasting and other predictive analytics play an important role. This also turns out to be a major differentiator for SAS relative to its competitors.

Business intelligence is not of necessity about tools and technologies; somewhat it is a plan of combining data from various sources with methodologies that make those facts harden in a unified manner.

   The data part of this strategy is data warehousing (see below Figure ). Once the data is sourced, scrubbed, enriched, conformed, and in conclusion housed in “access-ready” formats BI tools can make the data sing and dance.

   In above  Figure (adapted from Inmon, Imhoff and Sousa, 2002), BI is traditionally found in the Data release area (to the right side of the Information Factory.) Operational BI, on the other hand, is coverage and analysis done directly off of the operational systems or via an Operational Data Store (for example see, Imhoff, 2001.)

   Traditional BI makes use of past data points (what you know about the data from a historical viewpoint) and displays it for the end user to make main inferences. The historical reporting takes advantage of the dimensionality in the data to “slice and dice” by reporting facts along any number of dimensions.

   Early reporting tools allowable programmers to define exactly what they wanted to present in unreliable levels of granularity and aggregation. In the 1980’s a plethora of OLAP style data structures emerged, which included MOLAP, ROLAP and Hybrid-ROLAP. All of which provided the ability to drill in, around and through to make sense of the data presented.

   While OLAP is certainly not dead, highly structured interfaces to the data came out of an organization’s executive branch interested in the details. In other words, taking data from “green bar” and simply transferring it to the “browser” was not sufficient.

   Management required to manufacture the data into meaningful bits of information. “Tell me what’s wrong. Highlight the facts for me,” was the driving force at the back the dashboard and scorecards in today’s electronic toolbox.

   coverage on the past can only show what has happened, not what the future may bring. Past information must be joint with some real-time information and then covered with analytics in order to have true foreknowledge. This is where data mining, forecasting and other predictive analytics play an important role. This also turns out to be a major differentiator for SAS relative to its competitors.