The Next IT FrontierBy Corinne Bernstein | Posted 2009-03-03 Email Print
Better decision making is one of the last—and most significant—frontiers in business performance management.
Although most managers readily acknowledge the importance of improving the decision-making process, the topic is not yet top of mind at most companies, according to a new study, “Linking Decisions and Information for Organizational Performance,” conducted by Thomas Davenport, who holds the President’s Chair in Information Technology Management at Babson College.
Most organizations do not have specific agendas for connecting information and decisions, and are just beginning to focus on decision making and better use of information in the decision processes, according to the IBM-sponsored report.
“Decision making needs to take center stage—both from an IT standpoint and in general,” says Davenport. “Too many bad decisions have been made in both the public and private sectors of late.”
In the next two to three years, Davenport expects to see improvements in decision making and how companies use information as part of the process.
Since most large companies already have the basic application functionality they need and their transactions are largely automated, what remains to be done is to use the information from transaction systems to optimize business performance, he says.
Many of the technological components for better decision making are available today: These include data warehouses, business intelligence tools, workflow systems and decision rule engines. However, these tools are not well-integrated, and businesses are often not sure how they fit together, Davenport says.
Three Relationship Strategies
The study revealed three main levels of the relationship between information and decision making: loosely coupled, structured and automated. The most common strategy is loosely coupling information and decisions. Enterprises that take this approach might make information widely accessible to analysts and decision makers, along with tools to manipulate and display the information.
A second approach focuses on structured human decision environments. People still make the decisions, but efforts have been made to determine the specific information and other process resources needed to improve decision methods. These efforts create a stronger link between information and decisions. However, according to the report, this approach is not applicable for all decisions because of its narrower scope.
At the high end is the automated decision-making approach, which creates the closest connections between information and decisions. Human experts design the system, but they are not the primary decision makers. Automated decision making is most common in the financial sector—for instance, in program trading of equities, currencies and other financial assets, as well as in credit approvals.
The decision process must be well-structured and reducible to rules or algorithms, and these must be reviewed often to ensure that they continue to produce the right decisions. Since automated decision systems are costly and time-consuming to develop, companies should be selective in deciding which decisions to automate.
Regardless of which approach enterprises adopt, they must first ensure that any decision selected for “intervention” be important to an organization’s strategy and performance, the report states. Once this foundation is built, companies can concentrate on creating better information. More advanced steps include consciously addressing one decision-oriented process, focusing on improving decisions and establishing a formal decision management organization.
“Organizations have too much at stake to continue with the poor decision processes of the past,” the report states. “It is time for them to address better decision making as one of the last—and most important—frontiers of business performance management.”