Only 22% of survey respondents grade their company’s BI program as mature or established.  That means 78% of us could use some improvement!  What does Baseline consider to be the top three attributes of a mature BI program?

Several factors determine BI maturity, but ultimate success boils down to the degree of sophistication with which a company is able to leverage analytics to effectively make decisions and take action to run the business.  While BI maturity varies from industry to industry, and even department to department within a company, Baseline calls out three key attributes of a mature BI program:

• The data warehouse contains all relevant data—as determined through bona fide business requirements gathering—to meet users’ metrics for timeliness, usefulness, quality, and relevance to the business question at hand.  However, it’s not just about replicating volumes of detailed data.  In a mature BI environment, the data is modeled, managed, and integrated.  Users do not have to spend lots of time searching for the data or validating its meaning, lineage, or accuracy.
  
• End-users are self-sufficient.  They have data access “on demand.”  This means that business users can access metadata, understand the meaning and differences of the data, and launch reports, run queries, and perform complex analysis without involving technical staff.  The focus is on the end users and how to support their vastly different analytic needs, skills, toolsets, and data access rights. 
  
• Management understands the value of data and has established a culture of “fact-based” decision making.  The data is readily accessible to answer emerging business questions and fulfill new business needs in short order.  BI is considered fundamental to strategic business initiatives, and it is budgeted and staffed accordingly.

For BI users, real-time BI enables moment by moment business analysis. It involves using incoming transactional data to analyze the whole picture, detect unusual patterns, and take immediate action.  Real time BI is also known as “right time” BI, since the degree of data latency is informed by business need.

For example: A computer manufacturer may run multiple assembly lines using identical components on each line, but from different suppliers.  One supplier may have provided a bad component lot. While each assembly line builds the product, no errors are detected. But, as products come off the assembly lines, they are tested and yield a major spike in failures. With real-time BI, the component lot for that one supplier can be isolated out of all the work in progress to prevent potential failures in future products.

In another example: Credit card issuers continuously monitor card usage to detect unusual purchasing patterns across cities. Fraud usually occurs in a short period of time across many merchants in multiple cities. By analyzing real-time purchasing patterns as they occur, credit card companies can identify fraudulent use and greatly reduce losses.

For BI developers, real-time BI changes the data loading process.  Traditional BI development involves loading data at a specific point in time, usually overnight. While this approach simplifies the development and operations processes, it becomes inflexible to support real-time data needs. The real-time approach of loading data becomes more of a trickle feed, allowing transactions to be inserted into the data warehouse immediately as they occur. Because this is transaction based loading, it’s important to ensure that the DBMS locking capabilities don’t impede on-line reporting access.

According to the survey, respondents want to shift the business users' and executives' view of BI from "a set of tools" to "a business solution". How does Baseline recommend doing this? Why is this shift in perception important?