What do you know? When sharing knowledge, relationships matter

March 04, 2013

Some workers turn to the Internet when they need answers to questions. Others pop down the hall in hopes that a more knowledgeable coworker is available to help. For managers who want to make such knowledge-sharing in their organizations as efficient as possible, the best strategy just might be a combination of all three -- technology, social networks and available time.

That’s one conclusion that Su Dong has drawn from his research on how workers in knowledge-intensive industries share what they know about the tasks they perform. Dong and fellow researchers say that taking a systematic approach to knowledge-sharing can make the process more efficient and increase value, resulting in greater profits for a company or higher customer satisfaction for an organization.

“I’m absolutely fascinated about the social network part and also about the task assignment part. But how to put them together? That’s how I went into this business,” said Dong, clinical assistant professor in the Department of Information Systems at the W. P. Carey School of Business.

Other researchers have looked at how organizations can use technology to share knowledge, but Dong thinks it is equally important to look at how workers’ social relationships affect who they turn to when they need to learn something.

His research focuses on knowledge management in knowledge-intensive industries, particularly those that deliver software or information technology as a service. His findings also can apply to open-source forums and online communities and to other knowledge-intensive fields such as consulting, financial services, health care and education.

Obstacles to sharing knowledge

In any organization where workers have different skills and different levels of skills, and where they need to share their expertise with each other, they typically run into three main obstacles, Dong says:

  • They don’t know who knows what.
  • They don’t know who knows whom.
  • The person who has the knowledge might not be readily available to share it.

For example, you might be working on a project and run into a problem involving the Java programming language. You might not realize that your new colleague across the hall is a Java expert. And you might not know that your office mate used to work with the expert and is willing to introduce the two of you. But even if you meet, can the Java expert make the time to share what you need to know to complete your project?

Organizations are addressing the “who-knows-what” obstacle by using technology from companies like IBM Corp. and Microsoft Corp. to document employees’ expertise. It can be as simple as posting their biographies and listing their skills on a company intranet, or linking to their wikis or blogs. Or documenting expertise can be as advanced as using software that monitors email traffic to discover who is up to speed on particular topics.

It is important to recognize that employees vary in their types and levels of skills, Dong says. Organizations typically have workers who can be classified into three skill levels -- expert, average and novice. Dong is looking at how the three classes share their knowledge.

Despite the recent popularity of technological databases, other research shows that most workers who need to learn something turn to their real-life colleagues, hoping the colleague knows the needed information or at least has a connection to someone else who knows it. This is where the “who-knows-who” obstacle comes in, and where Dong thinks organizations need to pay more attention to the social networking concept of ties. Ties are the relationships formed when workers interact with each other and exchange information, over and over again.

Coworkers form strong ties when they share a location or work together on a project, allowing them to interact frequently. Forming strong ties takes time and effort, but the payoff is workers who are more willing to help each other. Coworkers have weak ties to each other when one is a friend of a friend, or when two workers share a mutual acquaintance but do not know each other directly. Another form of a weak tie is the performative tie, in which one worker looks up published information about another worker, but the two do not know each other personally.

The third obstacle, the availability of a coworker to share his or her knowledge, can make or break the knowledge-sharing efficiency of an organization. The knowledge gap between an expert and a novice might be large, Dong notes, so it will take more time for the expert to transfer his knowledge to the novice. The extra time could hurt the organization because it is taking the expert away from his or her own higher-value projects. Instead, Dong’s research found that workers share knowledge most efficiently when they do it in short, frequent bursts, and when their skill levels are similar. That means it is more efficient for expert-skill workers to transfer their knowledge to average-skill workers, and for average-skill workers to transfer their knowledge to novice-skill workers.

For example, if you are a novice, you are more likely to be located near or be assigned to projects with average-skill workers who could answer quick questions about your basic Java problem. If you are an average-skill worker, you are more likely than a novice to be located near or have worked on projects with an expert who can answer your quick questions about a trickier Java problem.

The implications for managers seeking to efficiently spread knowledge clearly point to the important role average-skill workers play in an organization. Average-skill workers are best positioned to acquire information from experts quickly and efficiently, and best positioned to give information to novices quickly and efficiently. Because of these benefits, it makes more sense to forge strong ties between experts and average workers, and between average and novice workers, Dong says. The novices are still somewhat connected to the experts, he notes, because strong ties between expert and average workers give the novices indirect, or weak, ties to the experts.

Better ways to overcome the obstacles

Current research tends to look at each obstacle to knowledge-sharing separately, Dong says. Instead, he and his colleagues think that organizations should develop a knowledge-management solution by addressing all three.

“We are proposing that we should look into those three types of information when making informed decisions regarding which worker should be assigned to what task and what types of help sources we as an organization should provide to that worker in order to improve the organizational performance and to facilitate knowledge diffusion inside the organization,” Dong said.

The researchers have developed a mathematical model that includes variables such as workers’ skill levels, their ties to other workers, and their availability to share knowledge. Their model allows managers to experiment with many combinations of variables and determine the optimal combination for their organization, without the disruptions that experiments with the real workforce would entail. The model also provides a framework that managers can use as a benchmark.

“The art of using a mathematical model is you can experiment with different combinations,” Dong said. “This will help the organization or the decision-maker think about the values for the parameters, based on their specific, unique environment, and think about how to evolve their network structure or improve the performance of their existing structure.”

How can organizations apply Dong’s research? He recommends these four steps:

  • Step 1: Look at what knowledge the workers have. Profile their varying expertise, and classify their skill levels in each skill as expert, average or novice.
  • Step 2: Use your information on what workers know, and on who knows who, to improve the knowledge flow as well as assignment of tasks to workers. Identify workers who can transfer their specific knowledge to coworkers, identify the existing social networks, and identify the types of ties that are best to use or that need to be created.
  • Step 3: Understand how to actively improve the performance of an existing network structure. Assign training systematically, and consider the types of ties to use. Be aware of worker turnover and how losing those workers’ ties can disrupt the “knowledge flow network.”
  • Step 4: If you can change the structure, or network, of how workers relate to each other, decide which one to adopt. Random networks, where information is disseminated widely to random individuals, tended to be most efficient at sharing knowledge because of the many strong and weak ties formed between workers. But there are also clustered networks, where information flows mainly within cliques of similarly skilled workers and builds strong ties; and small-world networks, where information flows both within cliques and randomly across cliques.

The result of this research would be a systematic approach to designing and using a knowledge flow network that maximizes workers’ knowledge over time. Ideally, the network would maximize the level of knowledge in an organization and maximize the spread of knowledge throughout the organization. This also would minimize the variations in worker skills, allowing almost any worker to handle almost any task.

In the future, Dong said, research could look at ways for workers to form strong and weak ties and at finding the best structure for maximizing the knowledge in an organization. He is interested in looking at how an organization’s market responds to its improved knowledge levels.

For now, Dong says, more efficient knowledge-sharing could have measureable benefits to organizations.

“They’re going to have better financial performance and operational performance,” Dong said. “They can get to their customers faster and serve their customers better, and at the same time enjoy a better workforce and better revenue.”

Bottom Line:

  • Document who knows what in an organization.
  • Recognize the importance of social networks, or relationships between workers. The most knowledge is shared between average-skill workers and experts, because these two groups tend to have strong ties and the difference in their skill levels is small.
  • Minimizing the time it takes to transfer knowledge is part of the equation. It takes a worker time to share knowledge with another. Workers might have to wait get knowledge from others, and teaching takes an expert away from his or her own job.