Depend upon it there comes a time when for every addition of knowledge you forget something that you knew before. It is of the highest importance, therefore, not to have useless facts elbowing out the useful ones. -- Sir Arthur Conan Doyle (1887)The departmental head talks to his minister once a month. Everybody who works under the head suffers as the dread appointment looms because his quest for information transforms him from mild-mannered Dr Jekyll to ravening Mr Hyde. A bank's customer relationship strategy details how it should employ its customer knowledge, integrate its delivery channel and segment its customer base. But when you phone the call centre, walk into the branch, or use the Web site, it doesn't feel like the bank knows you at all. As the information age slowly yields to the knowledge era, the two situations above might well be put down to failures in knowledge management (KM). Because, surely, if knowledge management is ever to live up to its hype in most businesses, it will do so because of its ability to help organisations do at least two things they cannot do well today.
One, it will help them to know what they already know. Who knows what about our customers around here, and how can we put that knowledge to best use? And two, it will deliver to all who need it the knowledge they want, before they know they want it. This cyclic and predictive element, while largely missing from most knowledge management strategies today, is increasingly closer to becoming a reality as KM heads towards some kind of maturity.
The knowledge management industry has been without a defining principle for so long, says IDC (US) analyst Gerry Murray, that it has resulted in lack of focus, abuse of the term, inability to gain momentum, and significant market backlash. But start with the premise that every request for information is motivated by business conditions that demand supplemental information to be resolved, and the challenge for KM becomes a way to capture these underlying business conditions -- the context -- as unintrusively as possible. "Some IT environments -- such as workflow, call centre, and help desk -- already perform this function in a limited way," Murray says. "KM must go beyond specialty installations and provide a generalised capability across structured and unstructured environments, so that corporate expertise is not restricted by departmental or systemic boundaries." In Murray's schema meta data becomes the seat of knowledge about how organisations use information. By capturing business-value triggers organisations will finally begin to understand how their information inventories are relevant to different business contexts in the organisations. And there's a second half to the equation, where knowledge management really begins to hit pay dirt. "If we know some circumstances in which particular information objects are valuable and if we can anticipate or detect those circumstances developing elsewhere, then we can automatically deliver relevant information proactively."Murray sees the abilities to capture business meta data and to anticipate developing business conditions as close cousins, with much of the theoretical underpinning already proven by case-based reasoning technology and some AI applications. The difference between the latter and KM is that with KM all that is needed is a small window that pops up only when activated and displays a dynamic top-10 list of related information objects or information clusters (such as a stored query). "When we have a system that knows how to match the right information with the right person at the right time, we will have knowledge management. Yes, significant product development must take place before we get there, but we are not talking more than 18 months ahead of where the KM market stands today," Murray says. "The result of delivering on this vision is the elimination of intellectual rework, the single greatest source of inefficiency in organisational life today." In Murray's mind, the greatest benefit of KM will be when it comes to making those once-in-a-decade, $100 million strategic choices that organisations will have to live with for many years to come.
At Ernst & Young, where there has already been a four-year focused global effort to transform the company into a knowledge-sharing organisation, the goal of proactive knowledge delivery has been identified for years. When consulting principal Andrew Keene joined four years ago the company had a slogan: "We want to have the knowledge we want, before we know we want it." Now that vision is slowly becoming a reality. In businesses like Ernst & Young consultants have no way of knowing what clients will ask for until they make the request. If a client is to get a speedy answer, the material should ideally already be on hand and accessible. The answer lies in the organisation's knowledge framework.
"The knowledge framework is about the processes we use to ensure that we capture stuff out of everything we do," Keene says. "These are often very simple things: like the fact that, as part of our quality process, staff according to their terms of engagement have to suitably remove all confidential material and submit suitable material to our database. Then we have processes to take that stuff and categorise it by level of expertise." Global Subject Matter expert teams devoted to areas like customer connections or customer relationship management handle the categorisations and the placement of material into so-called knowledge objects repositories. Material they identify as useful is placed in a power pack and delivered to those likely to need it, through either Lotus Notes or intranet. The real thrust of the KM effort now, says Keene, is to target much more highly the stream of information flowing to consultants. "Joining Ernst & Young is a bit like drinking from a fire hose, because there is just so much information that you can get," he says. "What we're trying to do is put some nozzles on the fire hose. So if you are focused on a particular industry, let's send what we know will be useful and let's appoint a knowledge steward for your service line or your industry group or your account, whose job is to filter and funnel."According to Jeff Sussman, chief executive officer with Delphi Australia, offspring of the well-recognised Boston-based knowledge-management consultancy, "knowledge portals" will play a significant role in the survival of organisations in the new infocratic economy, where knowledge will be the competitive resource. In Sussman's view, the organisations that are not only coping but moving ahead of the pack under extreme competitive pressure are those that are conscious of the value of their people resources and know their core competitive competency resides in the collective wisdom of their people.
"Looking to the future, successful organisations will be those that build corporate portals that capture knowledge about their people, processes, products and customers. From this knowledge they will be able to manage connections, ensure that they manage their overabundance of knowledge as opposed to their overload of knowledge, and ensure that they are managing multiple product or service life cycles. That will put them [the organisations] into the position to deliver what Peter Drucker calls managed abandonment: where they are not at risk from having competition [make] obsolete their products," Sussman says. "It will not be easy, as there is little past experience that will be relevant to the future. But there will always be a class of managers who will be able to harness the survival instinct in their organisations; who will find ways to manage the delicate balance between the people resources and will clearly understand that technology on its own is a necessary but insufficient enabler." The technologies such organisations are concentrating on are those that help manage what Sussman calls "the P2P connections" -- people to people, people to processes, processes to products and products to people. The technologies involved are almost always browser-enabled to allow a single point of access and minimal cost and time for deployment.
The US Central Intelligence Agency (CIA), headquartered in Langley, Virginia, has to deal with a daily cascade of information from all sorts of media and a Babel Tower of languages. An in-house precision dissemination system, or "precision push", helps analysts convert that information into knowledge for sending up the line to the president, the cabinet, law enforcement agencies and war fighters at the Pentagon. The system involves electronic distribution -- a security enhancement over paper-based delivery -- human gatekeepers to disseminate the information to the right analysts, and strong authentication.
Agents of the CIA reportedly expect to see commercial versions of the system emerge over the next few years. In the meantime, according to HCi Consulting principal Onno van Ewyk, many organisational knowledge management efforts are emulating those developments occurring on the World Wide Web. The two key trends are using channels and adopting technologies such as XML (Extensible Markup Language) that allow organisations to integrate information with more dynamic processes like transaction processing and even context-sensitive systems.
"Knowledge management is borrowing heavily from software development, to generalise these ideas about how people get things done so that they go beyond simply a particular application environment into a generalised work environment or a generalised organisational development environment," van Ewyk says.
Let's go back to that departmental head with his dire monthly Date with Destiny. The real issue here is the way good practice organisations get information, with value added to it, onto people's desks when they need it, says group head, human resources with PA Consulting Phil Probert. And the starting point is an organisation that is already good at asking itself the right strategic questions. "The clients I work with who don't have this kind of problem are the organisations that already have a good, live, robust process for discussing strategy issues. In other words, they're already accustomed to asking themselves, not just once a year during the strategic plan ritual but all the time, important strategic questions." Probert says he is working with one client at the moment that is superb at doing just that. Certainly, there is a formal strategic planning process, but at any one time that organisation has what it calls key strategy questions on the table -- what is likely to happen about X, or what would we do if Y happened. The executive management team keeps each such question alive until they are satisfied that they have worked out the scenarios and the answers. And they consciously and deliberately create a dossier of information around those questions that they syndicate amongst themselves.
"Somebody on the nine-man team owns each of the questions and acts like a market storeholder," Probert says. "Over the weeks and months he accumulates information which others can view. From time to time the CEO says: 'We are going to spend 20 minutes at the next committee meeting seeing where we are with this and I want to be reassured we are getting further with the answer'." Such organisations rarely get tripped up or spun into crisis mode, because they have identified the important strategic questions and are dealing with them as part of their routine. "I know their managers would say they have exactly the same problem as every other company; but I would rate those organisations as markedly better at continually looking at the curly questions and accumulating information and consciously discussing it," Probert says. PA Consulting accumulates the "good oil" its own consultants can access worldwide via public folders in its own database and internal networking systems, and uses the analogy of a marketplace. The market stores are the formal folders into which formal information is lodged, each has a keeper, and there's a wide range of intranet discussion groups. "The ownership of that information is quite important -- you can go and dip into it, but you are not allowed to alter it.
Only a small number of people are charged with and allowed to add to or subtract from and edit that data," Probert says.
Simsion Bowles senior consultant Mathew Geake sees at least three approaches an organisation could adopt to handle keeping its hypothetical minister up to date, or handling queries the hypothetical minister might fire at the chief executive. The first approach would be to develop a knowledge map. The organisation would ask everyone in the organisation to whom they talk when they want to know about a particular subject. The end result is a spider web of lines that focus on a few key people. "You find that maybe 5 per cent of the people in the organisation are actually the founts of all knowledge," Geake says. The second approach would be issues management. This would identify perhaps a dozen key subjects the minister is always asking about, then develop discussion groups or forums on each, ensuring the right people are involved.
"If you want to find out the current thinking within the organisation you go to a discussion group, look at the transcripts, talk to the convenor, and then you have got a very up-to-date picture of what sorts of discussions are happening in the organisation on that topic," Geake says. "That means you have to anticipate the broad categories that the minister is interested in, but you should be able to do that without too much difficulty."And third, one should look to the emerging notion of intelligent agents. These are little pieces of software that shoot around the WWW looking for information on particular topics to report back to whoever registered an interest in that topic. "So the intelligent agent is out there just keeping a eye on what is going on out there in the world, and what is coming up that might be relevant in that particular sphere. The technology is there, it is real and it can be done, but the uptake is taking time," Geake says.
A Knowing Consortium
While intelligent agents help find and filter information, customise information views and automate work, the Knowledge Sharing Effort is working to develop conventions facilitating sharing and reuse of knowledge bases and knowledge-based systems. The Knowledge Sharing Effort is a consortium of research centres including AT&T, Carnegie-Mellon University, EITech, Hewlett-Packard Research Center, Lockheed, Northwestern University, to name just a few. Building new knowledge-based systems today usually entails constructing new knowledge bases from scratch. The Knowledge Sharing Effort researchers are looking to ways it could instead be done by assembling together reusable components. The idea is that system developers would then only need to worry about creating the specialised knowledge and reasoners new to the specific task of their system, which would then interoperate with existing systems, using them to perform some of its reasoning. Another piece of the KM puzzle may well prove to be data visualisation technology. Amongst the first to deploy this in the 1980s was the CIA, with its "A picture is worth 1000 words" project, which created a new means for analysts to assimilate all of the documents, images and maps relevant to their missions -- what the intelligence community calls "unstructured data".
While such technologies will probably play crucial roles in the future of KM, analysts warn that one key to successful knowledge management is not to get hung up on the technology. If organisations spend so much time formally codifying knowledge that they forget the power of ad hoc conversational brainstorming, creativity is certain to be the ultimate victim. Sometimes, a 10-minute yarn around the tea urn can be as powerful a means of knowledge sharing as any intelligent agent or KM folder. Sometimes, the mere act of typing information into a database can rob it of the intuitive spark generated by face-to-face conversation. Ultimately, it seems culture is as much a key to KM as technology will ever be.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.