acm - an acm publication


A Ubiquity interview with Thomas H. Davenport

Ubiquity, Volume 2005 Issue September | BY Ubiquity staff 


Full citation in the ACM Digital Library

UBIQUITY: At the beginning of your book you've reprinted a cartoon showing an executive asking his assistant, "Wilson, what exactly is a knowledge worker and do we have any on the staff?" That's pretty funny. How realistic do you think it is in typical organizations?

DAVENPORT: Well, I certainly think there's a lot of fuzziness, ambiguity, and imprecision about what a knowledge worker is, and it's not a term most managers use easily. They don't say, "Okay, these are my knowledge workers, these are my non-knowledge workers." So despite the fact that the term's been around for a long time, very few people have been comfortable using it as a managerial concept.

UBIQUITY: So how would you define a knowledge worker?

DAVENPORT: I define them as people with high degrees of education or expertise whose primary job function involves some activity related to knowledge.

UBIQUITY: Let's do a pop quiz. What about an airline pilot?

DAVENPORT: I would say yes because pilots are applying a significant amount of knowledge as the primary part of their jobs.

UBIQUITY: What about actors?

DAVENPORT: Yes, because in interpreting the script their primary job is applying their acquired knowledge to a specific role.

UBIQUITY: So then you're saying that knowledge work comes in various types and is present in various degrees, is that right?

DAVENPORT: Yes. That why I emphasize that we're talking about one's primary job role. One of the confusing things about the term is that virtually everybody has to make some use of knowledge in a job. Making pizza requires some knowledge, but knowledge is not a primary component of the job.

UBIQUITY: In the case of people who think for a living, is there a significant distinction between people thinking new thoughts and people whose knowledge work is more tightly focused? In other words, is somebody in the planning department more of a knowledge worker than somebody in the accounting department?

DAVENPORT: Well, I have always maintained that they're both knowledge workers. They are different types of knowledge workers, but one is applying knowledge and one is creating knowledge. (And, you know, you have to be pretty clear that you don't want an accountant creating new knowledge. We tend to send accountants to jail for that sort of thing.) But I would define them both as knowledge workers.

UBIQUITY: You've been writing, consulting, and lecturing about "knowledge management" and "knowledge work" for some time now. How hard have you found it to get people to understand the concepts?

DAVENPORT: Well, I think knowledge management turned out to be a little bit of a detour in a way, since it became clear that most knowledge workers don't really have the time or the patience to sort through a big repository of knowledge to find what particular knowledge they needed. My new book is at the intersection of those two things, and combines a process perspective (in other words, how we improve work), with some of the ideas from knowledge management about making knowledge workers more effective and productive.

UBIQUITY: That cartoon we already talked about show's tells us what the executive asked his assistant but doesn't let us in on the assistant's answer. What do you suppose the assistant's answer would most likely be?

DAVENPORT: That's a good question. "Gee, boss, I don't know, I'll get right on it." But that would be pretty unlikely; maybe a better cartoon comeback would be, "Yes, boss, we've got 3,492 of them" — because nobody in real life would have that level of precision about who was a knowledge worker and who wasn't. In the United States we still have a kind of anti-intellectual bias. In some research my colleagues and I did a few years ago we found a curious reluctance by executives to even admit that some people use knowledge more than others or might be more valuable as a result. I thought that was strange, because obviously we have no problems admitting that executives get lots more money than anybody else, and allowed to have bigger offices with thicker carpeting and so on. Yet even though knowledge workers are for most organizations now creating most of the value in those organizations, we somehow don't want to admit that they exist. There's a kind of a funny political correctness at play in all this.

UBIQUITY: For awhile there was a lot of corporate interest in creating jobs with the title Chief Knowledge Officer. Where does that stand now?

DAVENPORT: Well, there were a flurry of them and then in the post-dot-com era a number of those positions went away. But the surprising thing about this whole knowledge management thing is that it refuses to go away altogether, and there are still some Chief Knowledge Officers working away happily. And there now seems to be a pretty widespread recognition that we need to make knowledge workers more productive and effective. But I don't really want to cast my new book that I've written as a knowledge management book; it's really more about how we can be more successful in designing the second generation of knowledge management.

UBIQUITY: So then what does it take to develop a good knowledge worker?

DAVENPORT: Well, the approach adopted by most organizations in the past is that all you really had to do is hire high-quality people in the first place — I refer to this as the HSPALTA syndrome ("Hire Smart People and Leave Them Alone). Of course, it's always a good idea to hire smart and capable people, and so it's always a good idea to put a lot of energy into attracting the right kind of knowledge workers and retaining them; but in my book I really focus on what are the kind of interventions you can make into making knowledge workers more effective. These include technological interventions, managerial interventions, workplace interventions, and social interventions.

UBIQUITY: Talk a little about those different kinds of interventions, starting with technological interventions.

DAVENPORT: I divided them into two types. First, there are things you do at an organizational level — things that from an IT perspective would be seen as applications for helping groups of knowledge workers do their work. One of my favorite examples is the approach taken by Partners HealthCare, which takes the knowledge a physician needs and embeds it into the structure of the work, and provides an electronic ordering system that has all the knowledge built into it; it's been fantastically successful, and basically every hospital in the land would like to do something similar. You know, it would be quite unsettling for a patient to hear the physician saying, "I've never really seen this before, I'm going to go do a quick Google search on it." For some reason we think that's bad behavior in physicians, though in other professions we think it would be very responsible. But with the Partners HealthCare approach the knowledge is in a sense injected into the nature of the daily work. In my book I give some other real-life examples but there aren't nearly as many of those as there ought to be.

UBIQUITY: Interesting. What's the second category?

DAVENPORT: And then the other, more general category I talk about is personal information and knowledge management, which is even less common than organizational applications. Every organization gives its knowledge workers Outlook or Lotus Notes or a spreadsheet or something like that, but they don't really help the knowledge workers figure out how to use these things effectively despite. The average knowledge worker spends 3 hours and 14 minutes a day just processing electronic information, so why not lend them a hand in doing this effectively? Intel is a particularly good example or an organization that does this right. Intel employees do 8,300 Web conferences and 19,000 audio conferences a week, so the company decided, Gee, if we're going to do these quite this frequently maybe we could help people by giving them some templates for running an effective meeting with electronic tools, and we could let them customize their office-related software to accommodate these things. So Intel did some nice segmentation based on how mobile the workers are and how aggressively they use new technology, and the company used the segmentation data to figure out what solutions made sense for what different types of knowledge workers. It's probably the single most aggressive firm on this kind of personal dimension.

UBIQUITY: Tell us now about the other types of intervention are.

DAVENPORT: First, there are interventions into the physical workplace. The frustrating thing about this one is that we do these interventions every day but we don't really learn anything because we don't do any measurements.

UBIQUITY: Example?

DAVENPORT: We move people from closed offices to open offices, or we move them into shared-office "hoteling" environments, or we tell them to work at home instead of coming into the office; in fact, we typically change multiple things at once, and we give them a new manager and new computer stuff and change the office kind of all at the same time. So we have not a clue about what's effective and what's not. A good example of creative workplace design is a place like IDEO, which basically offers free choice about what you want your workspace to look like. You're given your little area and if you want to bring in an airplane wing because you think it's got a really inspiring design and will help you be a better product designer, that's just fine with IDEO, as long as you can find a way to fix it to the ceiling so it won't fall on anybody. It uses a "mass-personalized" solution, which combines a group work setting low on segmentation (all employees share the same standard group work setting), but high on individual choice (employees are encouraged to bring in their own creative accoutrements to supplement the group solution).

UBIQUITY: Tell us now about the social networks dimension.

DAVENPORT: The social networks dimension is based on a hypothesis that I and a couple of other collaborators had; what we did is looked at high performers within four different organizations, and we found that they all appeared to be really good networkers, who used their networks more effectively than others did. They themselves were sought out for information, and in turn they knew where to go for information in the networks. When we interviewed them they talked about how important their social networks were to them and how they always tried to reciprocate when people sent them things. And so our hypothesis was that maybe one of the reasons they're high performers is that they're really good networkers, and that if we could make more people good networkers we'd have more high performers.

UBIQUITY: Let's pause a moment and ask you to talk a bit about yourself. You hold the President's Chair in Information Technology and Management, you're an Accenture Fellow, and you're the Director of Research at Babson's School of Executive Education. You also lead three sponsored research programs on knowledge management, process management and innovation. What can you tell us about all that?

DAVENPORT: I tend to go back and forth between various academically-oriented consulting jobs, and consulting-oriented academic jobs. I ran a research center for Accenture before I came to Babson. In the research centers I try to produce managerial knowledge that's actually useful — and unfortunately most business schools are not terribly interested in producing knowledge that's truly useful to practitioners doing real work. I've always found that the best way to do useful research is to get researchers in close proximity with business people who have real business problems and have found real business solutions.

UBIQUITY: What's Babson College like?

DAVENPORT: Babson is a business school, most known for entrepreneurship — it always dominates the entrepreneurship rankings. We've started a new research program on entrepreneurship in large organizations. Babson students are either undergraduate business students or MBA students, and there will soon be a few doctoral students in entrepreneurship. Many of the students are interested in either starting up their own businesses or being entrepreneurial within a larger organization. Babson is probably better known in places like Thailand and Chile than it is in the United States because it's highly attractive to rich Latin American and Asian families who want to train the next generation to take over this business. Babson turns out to be a pretty good place for that sort of thing. And certainly another nice thing about Babson is that it invites lots of business people into the classroom. It rejects this sort of invidious distinction between people who have Ph.D.s and do useless research versus everybody who does real work.

UBIQUITY: Well that's good.

DAVENPORT: Yes, I think it is.

UBIQUITY: What about the executive education?

DAVENPORT: Babson is also quite highly ranked in that regard, and we like to say that it's in the top 10 or sometimes the top nine institutions worldwide. It has a big executive education facility. We are known more for custom education than for so-called open enrollment programs where anybody can come and learn about finance and accounting for the non-financial manager or something like that. I think it's a nice mixture to do these sponsored research programs in the School of Executive Education, because we get a lot of spillover of content into executive education programs and the same kind of companies come through for research and education.

UBIQUITY: What's the main thing you want your students to understand?

DAVENPORT: I'm encouraging them to think a little bit more about how knowledge workers work. Management has always been about telling other people what to do, but today's knowledge workers don't like that so much — they just don't like to be told what to do. So it's a much harder thing to manage a knowledge worker than it is to manage somebody who doesn't think for a living.

UBIQUITY: Do you think there's any distinction along those lines between computer knowledge workers and other kinds of knowledge workers?

DAVENPORT: I think there are different levels of computer knowledge workers. For example, low-level programming is a very different job from high-level architectural design. So there's some segmentation even within that single category of programmer. The good news is that computer-oriented knowledge workers have been the focus of a lot of interventions in the past. The whole idea of measurement is probably more advanced in computer knowledge work than any other category because (although they're not great measures) you can measure lines of code or function points or something like that. There are also well-developed measures of the maturity of the systems development process. So one could argue that computer-based knowledge work is a bellwether for other types in the sense that there has been more emphasis on measuring and improving it than many other types.

UBIQUITY: Do you still do a lot of consulting?

DAVENPORT: I really do more management education and corporate education than I do consulting. I do some consulting, but consulting usually involves more time spent within an organization than I can often manage to come up with.

UBIQUITY: Well, if you did have the time and some company screamed to you, "We need help, and we want you to come in and help us think about what we're doing" — and you gave in to the entreaties. How would you likely start?

DAVENPORT: The first thing I would do is start to think about what are the different types of knowledge workers and start to narrow the focus a little bit by figuring out which ones are most important to the organization's mission and objective. So if it's a hospital it's pretty obvious that you'd maybe look at the doctors early on; if I were in a pharmaceutical firm it's probably the drug development scientists that I'd want to focus on. You've got to narrow your focus pretty early on and say who needs help the most if this organization is going to be more successful.

UBIQUITY: And then?

DAVENPORT: And then I would probably start to try to get a better sense of how the work is done now. One of the things I say in this book is that you have to observe people carefully, you have to hang out with them. It takes a lot of time, and that's one of the reasons I don't do much consulting these days, because I feel it's an insult to workers of any kind to come in and interview them for 15 minutes and say "Okay, I understand your jobs and I know how to redesign them." And it's particularly insulting to knowledge workers, who feel that they have a pretty good sense of why they do what they do.

UBIQUITY: How would you assess the morale of knowledge workers generally?

DAVENPORT: That's an interesting question. I believe it's at a turning point right now. I think the morale over the last couple of years has been pretty low, because it's been a period of cost-cutting and disloyalty from the corporations to the workers. And the labor market was bad so there wasn't much a knowledge worker could do about it. But this fall is likely to be one of the best periods for finding a new job or being offered a new job that we've had in the last five or six years, and I think knowledge workers are starting to get a little bit more respect. Organizations are starting to say, "Oh my God, how are we going to keep these people whom we have abused for the last five or six years? We've worked them too hard, we've given them the message that we don't really care about them, so now how do we hang on to them?" So I think it's going to be an interesting period. In the dot-com era there was a lot of experimentation relative to knowledge workers — there were free agents and there were people who were not in traditional job roles. More recently, we've seen this wild card of offshoring and outsourcing, and for the first time knowledge work can go elsewhere. That's not necessarily good for morale, but I'm hoping that it leads knowledge workers and their managers to realize that we had better get a lot more productive about this kind of work or it won't be our work anymore, it will go elsewhere. And that realization certainly adds a certain level of urgency to these issues.

For more information about "Thinking for a Living: How to Get Better Performance and Results from Knowledge Workers," visit


Leave this field empty