Thomas H. Davenport

Thomas H. Davenport holds the President’s Chair in Information Technology and Management at Babson College and is responsible for the overall management of the Institute for Process Management. He and Larry Prusak also manage the Working Knowledge program. His published works include Process Innovation, Thinking for a Living, Working Knowledge and What’s the Big Idea? with Prusak, The Attention Economy co-authored with John Beck, most recently Competing on Analytics co-authored with Jeanne G. Harris.

Note: This interview was conducted in 2007. Davenport is now in the process of completing another interview.

Morris: In your various books and articles, you offer excellent advice as to how to manage knowledge. Let’s begin with a more basic challenge, one which Carla O’Dell and Jackson Grayson examine in If Only We Knew What We Know: First determining what your information needs are and then what necessary knowledge already exists within an organization. Your views on that?

Davenport: I certainly agree that it’s better to start with the knowledge your organization possesses already. It’s somewhat surprising that relatively few organizations have done either of the two steps above. They don’t examine their strategies and decide what information and knowledge are critical to achieving them, and they don’t have a good inventory of what they know already. Those two steps in that order would be a great boon for knowledge management.

Morris: In What’s the Big Idea? you and Larry Prusak explain how ideas are linked to business success, who introduces ideas to organizations and how they do that, why “content counts,” where the best management ideas come from, how ideas interact with markets, where to find ideas most appropriate to a given organization and then how to sell them, and why idea-based leadership is essential to any organization’s success. These are admirable objectives. Why do so few organizations achieve them?

Davenport: People get very excited about business ideas, but they don’t manage them to fruition very well. There are a variety of problems in this regard. Most companies take on too many ideas at once. They don’t have any sort of process for monitoring how the idea is being implemented within the organization. GE is the primary exception. Under Jack Welch they developed a management system for making new business ideas a reality, and they were very disciplined about which ones they took on. Then there aren’t usually enough idea practitioners around to make it all work.

Morris: Also in What’s the Big Idea?, you and Prusak assert that “Idea-Friendly Culture” which (a) has open dialogue between and among all levels, (b) supports “boundarylessness” to maximize individual and collective intellect from both within and outside the organization, and finally, (c) encourages trust and responsibility which will “allow people to learn effectively from each other and provide motivation for putting ideas to work.” That said, what role could and should senior management have to expedite and support such initiatives?

Davenport: They control the organization’s resources, and each idea that an organization adopts consumes resources. So it’s very important that they decide which ideas enter the portfolio that the organization will try to implement. They’re really setting the idea strategy—“what ideas are we going to pursue?” They also have to put pressure on the organization to make things happen. At GE, Welch would call business unit managers who would be moving a little slowly on an idea, and say, “Why aren’t you doing more with digitization? This is really critical to our success and your long-term future here.” That’s obviously very powerful.

Morris: As you indicate in Thinking for a Living, what do you consider to be the appropriate relationship between knowledge workers and various technologies provided to them?

Davenport: In most cases thus far, knowledge workers have been the victim of the technologies. A lot of tools have been thrown at knowledge workers, and nobody’s given them much help in thinking about how it fits their jobs and their objectives. Now we all have a lot of technologies—laptops, desktops, PDAs, cell phones, pagers, etc.—but they don’t integrate very well, and everything is very fragmented. The result is that only a small percentage (less than 1% in my informal surveys) of knowledge workers feel that they are very good at managing their personal information and knowledge environments.

Morris: In the same book, you suggest that individual knowledge work improvement initiatives have two attributes. With regard to the first, why should they be focused on improving the performance of knowledge workers as individuals, not as members of a larger group

Davenport: I did some work with the Software Engineering Institute at Carnegie-Mellon. I realized that they had figured out something important. If you want to improve how an organization gets better at software development, you need to address the problem on multiple levels: the company, the team, and the individual. I think the same thing applies to knowledge worker productivity and effectiveness. We really haven’t done enough at any of these levels for knowledge work.

Morris: Why should individually oriented initiatives be directed at improving some skill or capability, rather than instituting a new process?

Davenport: I really believe in both. Again, we should be working on multiple levels. Of course, you can’t do that for every job. You have to pick one or a few knowledge work roles that are very critical to your organization’s success, and focus on those.

Morris: In Chapter 7 of Thinking for a Living, you pose a very important question: “What’s more important to improving knowledge worker performance: technological networks or human networks?” For those who have not as yet read your brilliant book, what is the gist of your response to that question?

Davenport: Well, I try to go with the data, and when we asked high-performing knowledge workers how they get the information they need to do their jobs, they generally said they got more useful information from their human networks than from technological ones. Consider the implications of that.

Of course, you don’t really have to choose—you can try to improve both types of networks. The problem is that most organizations spend a lot more time and money on the technological networks, and ignore the human ones altogether. Knowledge workers are well aware of that neglect…and resent it.

Morris: In your opinion, what are some of the most common misconceptions about the practices of high performance knowledge workers and how they get their work done?

Davenport: I guess the biggest misconception is that you can’t do anything with knowledge workers, as I suggested above. Organizations just leave them alone. I think it’s possible to impose a bit more structure on knowledge work and measure and improve it in almost every case. Of course, you can go too far, and alienate these unique and very valuable workers.

Morris: Before concluding this interview, please tell us about your next book, Strategic Management in the Innovation Economy, which you co-authored with Marius Leibold and Sven Voelpel.

Davenport: That book is really a textbook about how to pursue multiple forms of innovation in the contemporary world. I don’t generally do textbooks, but Marius and Sven did most of the work. Right now I am most excited about my next book, which is about how companies compete on their analytical capabilities. It should be out in early 2007. I’m not a terribly quantitative person myself, but I see the world moving in that direction. I wrote an article on this topic entitled “Competing on Analytics” in the January (2006) issue of Harvard Business Review, and it’s gotten a more positive reaction than just about anything else I’ve written. So I’m excited about it.