Dan Power, Editor of DSSResources.com, conducted an email interview with Tom Davenport in May 2007.
Q1: How do you define the concept "Decision Support System?"
Davenport's Response: Well, I don’t really use the term much anymore, since most businesspeople now use the term “business intelligence.” But in any case I would define it as IT applications that help organizations to make decisions. I think it’s only the “support” word that is becoming obsolete, as systems are increasingly capable of actually making automated decisions, as opposed to just supporting humans in doing so.
“Business intelligence” usually refers to the use of technology for reporting and data access, as well as analytical applications. My research is more focused on the analytics side.
Q2: How did you get interested in DSS and Knowledge Management?
Davenport's Response: I got interested in knowledge management in the early 90s. As a sociologist I was interested in the human role in managing and processing information, and I was most focused on information with a high proportion of human contribution and insight. At some point I began to call that “knowledge” and began to write about knowledge management.
I had worked for a consulting firm that had done DSS in the 80s, but I didn’t do much research in it. I got interested in doing research on analytics and business intelligence in the mid to late 90s, largely as an outgrowth of my interest in knowledge management. I viewed analytics as simply a form of knowledge derived from data. Whereas in knowledge management there was a pretty high degree of orientation to the human role, there wasn’t much in analytics or data mining. The feeling at the time was that you didn’t really need much in the way of human skills—technology was going to find the relationships in the data and figure out the best models. But when several colleagues and I did a research project in 1999 looking at companies that were good at analytics, we found that they all tended to have lots of smart and quantitatively-focused humans around. My more recent work on analytics—culminating in my co-authorship of the book Competing on Analytics—continued to focus on the human factors in analytical leadership.
Q3: You were a leader in promoting Knowledge Management. What has happened to KM and especially Knowledge Management Systems?
Davenport's Response: It is certainly still around and to my mind has become a standard business approach for knowledge-oriented companies. A recent Bain and Co. survey of management tools and techniques found that it was in the top ten tools that organizations around the world use. Of course, it’s always evolving. I think that now organizations are incorporating knowledge into broader approaches to information and content management. And there are now some interesting new approaches to participative knowledge management like wikis and internal blogs. I don’t think they’re going to revolutionize organizations, but they will almost certainly have an evolutionary effect.
Q4: What keeps you interested in information systems?
Davenport's Response: It’s an ever-changing field, as the changes in knowledge management above illustrate. And I have always thought there is a good niche for someone like myself who focuses on the human side of computing and information management. It’s really the “sociology of information,” which didn’t exist when I got my Ph.D., but is beginning to be recognized as a valid field of sociological study.
Q5: What do we really know about how computerized decision support impacts decision making? What generalizations are supported by research?
Davenport's Response: We don’t know much. James March told us, for example, that managers who gather lots of structured information don’t necessarily use it to actually make their decisions. A couple of other researchers (Bill Bruns and Sharon McKinnon) told us that managers prefer to get information from human sources when they can. There really isn’t a lot of research out there about how managers make decisions, and there is even less about how they use computers. And that, of course, assumes a traditional “decision support” environment in which a human decision is informed by computer-based analysis. As I mentioned earlier, there are many alternatives to that traditional approach today. Obviously much more research is needed.
Q6: Do we need a theory of computerized decision support and knowledge management systems? If so, what are the main elements or a starting point?
Davenport's Response: I am not big on grand theory. I do support the development of “theories of the middle range,” as Robert Merton called them. They would include some general principles like: Decision-making is largely a human endeavor, so any focus on DSS and business intelligence should include a focus on how humans actually use systems to make decisions; Knowledge is both created and applied in the mind of a human knower, so attempts to manage it should deal explicitly with those humans.
Q7: How is knowledge management related to DSS?
Davenport's Response: As I mentioned above, I think they both are heavily reliant on human capabilities. I also believe that a finding or result from an analytical system is no different from any other piece of knowledge. It needs to be captured, shared, and managed over time. A few of the highly analytical companies I have researched are beginning to think about how they capture their findings in knowledge management systems.
Q8: What is the role of DSS in attention management?
Davenport's Response: Analytical systems, and even reporting systems, usually involve some sort of data reduction, so they allow firms to use less human attention to find out what’s going on. Systems that provide alerts, rather than regular consultation, are even better from an attention management standpoint. It’s also important when designing any of these systems to think about how the information can be made attention-getting. The “red light, yellow light, green light” displays are one simple example.
Q9: How can managers get more "usable" data for decision support?
Davenport's Response: Of course, having good data is the number one prerequisite for DSS or business intelligence. I generally encourage companies to engage their senior executives in discussions about what information should be defined and managed at the enterprise level, versus information that can vary in definition and format across the organization. This is a “federalist” approach to information management, rather than the feudalism or anarchy that many organizations have. Otherwise, I don’t think my advice is very distinctive—I believe in warehouses, marts, etc., like everyone else these days.
Q10: What do you see as major trends in computerized decision support? Where are we headed?
Davenport's Response: In addition to increasing use of analytics for prediction and optimization, we are also moving toward more real-time and embedded decision systems, and we’re already seeing that technology being incorporated in ERP and other transaction environments. We will see greater use of textual and even voice data. We’ll see a merging of analytics and reporting systems into “predictive reporting” capabilities. As these analytics proliferate in firms, we’ll have to have more model management repositories and systems. Of course, we’ll have to have more and more analytical people in organizations to make effective use of these capabilities. It should be a good time for anyone who teaches this sort of thing!
About Tom Davenport
Tom Davenport holds the President's Chair in the Information Technology Management Division at Babson College and is Director of Research for the School of Executive Education (SEE) at Babson. At SEE he is the Academic Director of the "Institute for Process Management" and "Working Knowledge" research programs. He is an Accenture Fellow and has taught at Harvard Business School, University of Chicago, Dartmouth's Tuck School of Business, and University of Texas at Austin. He has directed research centers at Ernst & Young, McKinsey & Company, and CSC Index. Professor Davenport wrote, co-authored or edited ten books, including the first books on business process reengineering, knowledge management, and the business use of enterprise systems. He has written hundreds of articles and columns for such publications as Harvard Business Review, Sloan Management Review, California Management Review, Financial Times, Information Week, CIO and many others. His book, What’s the Big Idea: Creating and Capitalizing on the Best Management Thinking, was named one of the three best books of the Spring 2003 season by Fortune magazine. In 2003, he was named one of the top 25 consultants in the world by Consulting magazine. Professor Davenport's areas of expertise are in Knowledge Management and knowledge worker productivity, Enterprise Systems, and Process Management. His M.A. and Ph.D. are from Harvard University. His website is URL http://www.tomdavenport.com/.
Power, D., "Tom Davenport Interview: Competing on Analytics", DSSResources.COM, 05/27/2007.
Check Tom Davenport and Jeanne G. Harris's book, Competing on Analytics: The New Science of Winning, Harvard Business School Press, 2007. Davenport and Harris argue that the frontier for using data to make decisions has shifted dramatically. Certain high-performing enterprises are now building their competitive strategies around data-driven insights that in turn generate impressive business results. Their secret weapon? Analytics! They recommend using sophisticated quantitative and statistical analysis and predictive modeling to improve decision making.