Knowledge Networks

Mapping and Measuring Knowledge Creation, Re-Use, and Flow

by Valdis Krebs

No one doubts that better management of knowledge within the firm will lead to improved innovation and competitive advantage. Everyone agrees on the goal -- better utilization of internal and external knowledge. It is the approach to this goal that is hotly debated. Many vendors and consultants push a technology-driven approach. "Buy our state-of-art knowledge storage system and you will never again lose knowledge that is vital to the company!", they exclaim. Other consultants emphasize the soft-side of Knowledge Management. "Create a learning culture, that rewards sharing, and knowledge management will take care of itself!", they postulate. There are no silver bullets. Not from the technologists. Not from the culture prophets.

The effective utilization of knowledge and learning requires both culture and technology. Explicit information and data can be easily codified, written down, and stored in a data base. For this type of business information we have the necessary skills and more than adequate tools. Yet, simple data is frequently not where competitive advantage is found. An organization's real edge in the marketplace is often found in complex, context-sensitive, knowledge which is difficult, if not often impossible to codify and store in ones and zeroes. This core knowledge is found in individuals, communities of interest and their connections. An organization's data is found in its computer systems, but a company's intelligence is found in its biological and social systems. Computer networks must support the people networks in today's fluid and adaptive organizations -- not the other way around.

Visualizing Knowledge Networks

The organization chart has been a staple in the Human Resource [HR] department. It displays who works where and who reports to whom. This was sufficient knowledge in a time when organizations faced gradual change. These charts where tools for control and planning. Today's fluid business environment does not allow only static structures and does not reward those that follow prescribed configurations in the face of rapid change. The fast economy requires flexible, adaptive structures that self-organize internally in response to changes externally. In this knowledge-critical economy we need charts to show us who knows what and as a complement who knows who. In addition to pictures of hierarchy we need visualizations of the massive interconnectivity that occurs in the learning systems that are our organizations.

Organizational Network Analysis [ONA] is a software supported methodology that reveals the real workings of an organization. It uses the rigor of systems analysis to reveal the behavior inside and between organizations. Knowledge networks are mapped that uncover interactions within and across the boundaries of the organization. These visualizations are in effect business x-rays of how things actually get done -- evidence of adaptation in the organization. HR Managers and consultants use these revealing diagrams in the same way that doctors use x-rays and CAT scans -- to see what is normally invisible. ONA exhibits both how knowledge is shared in emergent communities of practice, and how it is utilized in key business processes. In short, it uncovers the hidden dynamics that support learning and adaptation in the modern organization.

Not only can HR mangers and consultants now visualize the connections that matter, they can also measure and benchmark them. Based on recent research, an organization can now be benchmarked in key dynamics such as adaptability, capacity to learn, openness to the environment, ability to span boundaries, brittleness of its structures, probability of project success, and efficiency of information flow. This technology provides the ability to drill down into a complex organizational system and find emergent experts, opinion leaders, bottlenecks, breakdowns in communication and communities of practice. The organization can be viewed and measured from the system-wide level, to the group level, and down to the individual -- you can see the forest and the trees... and how they are related.

ONA is an outgrowth of many knowledge disciplines including social network theory, organizational behavior, interpersonal communications, chaos theory, complex adaptive systems, artificial intelligence-based search and pattern-matching, communities of practice research and a branch of mathematics called Graph Theory. ONA is basically an Object-Oriented model of an organization with objects such as people, teams, and technologies interlinked sending messages to each other and invoking their respective methods to accomplish the goals of the firm.

Organizational Layout

The organizational example described below is a combination of several knowledge management projects performed by the author. A key business process will be mapped along with the knowledge exchanges that support it. The organization will be viewed from several perspectives. First, the company can be viewed via prescribed structures such as hierarchy. This view reveals who is assigned where and who reports to whom. Next, the company can also be viewed via emergent structures. These views reveal what happens in the white space [between the boxes] on the organization chart. The emergent views also show where certain knowledge is clustered in the organization.

The model organization in Figure 1 below is divided into four components:

The Corporate HR office is divided into the 3 departments that participate in a critical HR process. Five key knowledge areas that contribute to this process where uncovered from interviews with the client's employees. Employees names are replaced by numbers to maintain privacy of the study participants. 

Figure 1 - Organization Layout

The first question that employees where asked was: "With whom do you exchange information, documents, and other resources in order to perform your role in HR business process X?" Below is a map of the work exchanges to execute this critical HR business process. These are all confirmed two-way [give and receive] interactions. The line thickness denotes intensity of relationship.

Figure 2 - Work Flow Network

Work Flow Network

The formal organization structure supports the work flow for this business process -- most of the strong work relationships are within the functional walls of the prescribed organization. Compensation & Benefits and HR Policy are strongly interconnected and appear to be working as one unit in this process. The SBU's HR offices do not work with each other directly. Most of their interaction is with the corporate HR office. This revelation alarmed the Executive VP of Human Resources. All SBUs have similar missions and very similar employee populations -- they should be talking to each other about the changes in this key HR program. As a result of this finding the most central node in each SBU was invited to process change meetings together with the other SBUs so that knowledge and experience sharing relationships would start to develop.

The knowledge exchanges around this business process are mapped next. These links reveal who helps who learn and make sense of what is happening in this business process. This is a map of how expertise is shared. Nodes that are central in this network are the experts that are sought out for critical information and knowledge to complete this business process. Which nodes appear to be 'in the thick of things' in the knowledge network in Figure 3 below?  How does the work flow network compare with the knowledge exchange network?

Figure 3 - Knowledge Exchange Network

Knowledge Exchange Network

Figure 3 reveals more inter-group connections -- knowledge necessary for this process is distributed throughout the organization. A greater number of links between the SBUs are discovered. Yet, corporate seems to hold most of the knowledge to execute this process. R&D has fewer connections within the corporate office and is now well connected to SBU 3 whose HR programs are holdovers from its former parent company before it was acquired. They apparently need more interaction to adapt to this new program.

A cluster discovery algorithm is applied to the network data to see if this knowledge resides in emergent communities of knowledge [aka communities of practice]. Communities naturally self-organize naturally in companies around common problems, interests, customers, and complex knowledge areas. It is within these communities where core competencies of organizations are stored, shared, nurtured and enhanced. Individual learning is enhanced by being a member of one or more communities of practice.

Emergent communities have formed around the 5 knowledge areas. They are mapped in Figure 4 below. To identify who is from which organization the reader can refer back to Figure 3 to see which node color corresponds to which business unit. Employees from SBU 1 are connected to each other in Knowledge Community C and E but are not tied to community members from other organizations. Community fragmentation like this is found in both forming, and dissolving, communities. The communities in this organization where just forming in response to a changed environment and new direction from the HR VP.

Figure 4 - Emergent Communities of Knowledge

Visualizations, like in Figures 1, 2, 3 and 4 above, give insight into complex human systems not readily available by other means. Even deeper insights can be gained from measuring these complex human structures. Networks can be measured on the individual, group, and system-wide basis. The focus here will be on individual network centrality. This measure reveals which employees are key in the flow of information and exchange of knowledge. A central node is in the thick of things and has access to diverse network resources such as knowledge, support, and other hidden assets in the organization. Employees with high network centrality scores have a greater capacity to get things done.

Network Connectivity

A common belief is that high network activity brings increased network benefits. This is not necessarily true. High network centrality does bring network benefits. Research has shown that employees who are central in key networks learn faster, perform better, and are more committed to the organization. These employees are also less likely to turn over. On the other hand employees with low centrality, those who are on the periphery, are much more likely to leave the organization. Project teams also benefit from being central in advice and expertise networks. Teams that are central in the organization's knowledge networks complete their tasks quicker than project teams who struggle to access the knowledge they need to perform their work.

The secret to network benefits is in the pattern of direct and indirect connections surrounding a node. It is the pattern of relationships, that a node is embedded in, that either constrain or enhance the ability to get things done in the organization. The goal is to obtain wide network reach without having too many direct ties. It is the indirect ties that provide network benefits. Research has shown that both individuals and groups who are central in organizational networks, yet are not overwhelmed by direct ties, are very effective in getting things done. Those burdened with too many direct ties are not as effective.

Opportunities in Networks

Innovation happens, within and between organizations, at the intersection of diverse information flows and knowledge exchanges. The network in Figure 4 above shows many opportunities to cross-fertilize knowledge -- connect knowledge communities that are not yet connected. These potential connections in networks are called structural holes. It is across these holes in the network that the opportunity-seeking player [individual, team, or organization] can establish a superior position where diverse knowledge and ideas intersect. This position is superior because it is unique -- these flows do not intersect anywhere else in the network. The node that spans the right structural holes receives a diverse combination of information and knowledge available to no one else in the network. From this advantageous position innovative products and services can be created. An organization whose employees effectively span these internal holes of opportunity creates a competitive advantage that can not be easily duplicated by competitors. Even if competitors hire away a few employees in the network they still cannot easily duplicate the knowledge community [the unique pattern of interconnections] that is in place in the other organization.

Possible New Knowledge Exchanges

How should these knowledge communities be connected? Use ties that already exist between groups -- the work ties that currently exchange task resources but not knowledge and learning. Find nodes that are not overloaded in the work network and assign them the addition duty of knowledge exchange. The links in Figure 5 below reveal who has a work tie, but not a knowledge exchange. It is these single purpose ties that can be expanded for multiple duty.

Figure 5 - New Knowledge Exchanges

Why did the HR VP look for possible connections in the emergent organization? Why didn't she just assign various employees to these boundary-spanning roles? She knew how emergent communities work -- trying to formalize the informal, or trying to steer an emergent process, just leads to resistance and then disappointment. Knowledge-based organizations, through the people in them, attempt to adapt to their environments. Exerting too much control over this process hinders effective outcomes. Building emergent communities and informal networks is a lot like gardening. The manager/gardener must provide resources and remove obstacles/weeds so that the employees/plants can follow goals/sunlight to self-organize and grow. Trying to exert too much control over this emergent process will usually result in a poor harvest.

Network Management

Once the people networks are understood, the right technology can be implemented to support these evolving entities. Computer technology needs to be as flexible as the adaptive, self-organizing human networks it supports. To meet this demand for adaptive technology many organizations are utilizing the flexible technology and protocols of the Internet inside the organization.

Tools to manage computer networks have been in existence for a few years and are becoming more sophisticated. Tools for human networks are just starting to emerge into general business use. ONA tools such as InFlow [used in this article] are aimed at HR experts and management consultants. ONA software has been utilized by early adopter firms since the late 1980s and is now gaining interest in many industries.

Network Models: Tools for the Connected Economy

Today's fast and fluid business environment requires HR mangers and consultants to understand the constantly changing economic webs within and between organizations. Static, hierarchical structures are no longer sufficient to function in the connected economy. Adaptation and Learning are joining Control and Planning in the executive suite of today's innovative corporations.

ONA has been used in many progressive firms including Rubbermaid, TRW, IBM, and Lucent Technologies. These firms have applied ONA to improving knowledge exchange, workforce diversity analysis, post-merger integration, process improvement, and organizational redesign. Consulting firms such as Ernst & Young LLP and the IBM Consulting Group have, between them, utilized this technology with hundreds of clients to support various projects such as product development, computer system implementation, organizational design, business transformation, retention analysis, business process reengineering, knowledge management, strategic planning and other organization effectiveness efforts. ONA has also been applied to understanding the emergent dynamics in the network of alliances between firms in the Internet industry.

A network view of the world is necessary to adapt to the chaos and complexity of continuous change. In the past, HR departments focused on the nodes [employees] in the network which were often modeled as boxes on a hierarchical chart. In times of reorganization the boxes and their formal connections where moved around by management prescription.

In today's fluid economy, HR must also focus on the ties [flows, relationships] in the network, and their ever-changing patterns. Many adaptations to the rapidly changing environment today are soft reorganizations -- knowledge exchanges and information flows are created/strengthened/weakened, but the formal hierarchy remains in place. This allows for more frequent and rapid adaptation. Obviously technology must be adaptable with these frequent soft reorganizations. Network models of how organizations get things done are as necessary in the new economy as organizational charts where in the industrial era.

Software and Training in social network analysis are available from the author.

Copyright © 1998, Valdis Krebs