Continue in 2 seconds

Making Business Intelligence Actionable

  • November 01 2001, 1:00am EST

Top executives are often not convinced that business intelligence (BI) is a must-have capability that directly impacts their businesses. The problem is that most executives believe BI stops with pixels on the screen or with beeps on the pager. Further, many BI professionals believe that the BI job is finished when the right information is delivered to the right person at the right time – or some similar paraphrase.

The fallacy is that information itself is pure overhead, delivering no value to the business. Information can only benefit the business when some action changes the course of some business process. To have value, information must improve your products, enhance interactions with customers or have similar impacts.

The issue is making BI actionable. The implication is that the BI must change the course of action within the enterprise.

Making BI actionable is essential for top management to view BI as a must-have capability. This issue has not received sufficient attention across the IT industry, although our colleagues in competitive intelligence have stressed this issue for many years.

Let me suggest a new approach for evaluating BI in terms of complete BI, which consists of the following functions: observe (What has happened?), understand (Why has it happened?), predict (What will happen?), react (What should we do now?) and reorganize (How can we do it better?).

Figure 1: Complete Business Intelligence

The observe function captures the history of the business. As the bread-and-butter of data warehousing, the flow from the operational transaction system to the warehouse environment has dominated BI efforts. As justified by the business, current challenges are lowering aggregation levels to that of atomic transactions and reducing the time from transaction commit to data available in the warehouse.

The understand function understands the dynamics of the business, as inferred from the data. Using OLAP and data mining techniques, the critical business dimensions are defined and analyzed. Designing the proper star schema is defining the primary dimensions to be analyzed, and time has always been a fundamental dimension. However, the explicit modeling of those business dynamics is lagging.

The predict function forecasts the future state of key business variables, based on the business model. Most BI efforts have ignored prediction, assuming that the past is always the best predictor of the future. Predictive techniques, such as neural networks, should be utilized generally, rather than confined to point solutions.

The react function decides and executes a course of action based on an understanding of the business dynamics and predictions about future trends. This is where the "rubber meets the road" in making BI actionable. In the past, we referred to this function as closing the loop from warehouses to legacy transaction processing systems. In current architectures, those boundaries are blurred with rich analytics embedded throughout the architecture.

The reorganize function learns to improve business processes by refining best practices. Improving a decision is not sufficient; improving the decision process is the challenge. The analogy is catching a fish for a person or teaching that person to fish. The latter is the focus of this function. The enterprise that continually improves (whether incrementally or radically) its business processes will be the enterprise that succeeds. The challenge is how to rapidly adopt and encapsulate best business practices within your business culture.

In summary, if your BI projects are not changing the way that you do business, then they should not be considered BI.1 Hence, you become the weakest link. Good-bye!

Evaluate your current and planned BI projects in terms of the five functions for complete BI. In particular, question the degree to which the BI project will improve the ability of your business to react to market changes and to reorganize in response to those changes.


1. Hackathorn, Richard. "Little BI Versus Big BI." DM Review January 2001.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access