How Social Network Analysis Works

Social Network Analysis is a statistical method used to analyze organizational structures. SNA provides a visual and mathematical analysis of human relationships that can be used to evaluate communication patterns and find holes in an organization's communication network. The tool uses people's personal connections to compute an organizational communication network and to map information flow between people and departments.

"When I saw the SNA output I realized I had found a new way to quantify the relationships within and across my teams. It let me see those who consistently communicate well with others," says Randle.

The first step in Randle's Social Network Analysis was to survey everyone in his organization. He asked only four questions:

• Who do you go to for advice or information when assessing a difficult problem or discussing ideas?
• Who do you depend on to get your job done?
• Who do you communicate with most frequently? and
• Who is your most valued contact within XO IT Operations?

Each person was allowed to name up to three people from within the IT Operations department in response to each question. The answers to the survey were used to generate a set of network diagrams called sociograms similar in form to a router network diagram. In a sociogram, the network nodes represent people, and the connections between nodes indicate communication between individuals.

One sociogram showed nodes that were sized according to the number of ties connecting an individual. "It was immediately apparent which people were central to my organization and who had many ties to others on my team," says Randle. "This showed me those who were strong communicators and sharers."

In social network analysis, what Randle saw intuitively is called centrality-a measurement used to define the person's relative importance to an organization based on his location in the network. Centrality takes different forms. Degree centrality counts the number of ties a node has to others and is used to measure a person's overall activity in the network. Betweeness centrality shows the degree to which a person connects different groups, thus controlling information flow across the network. Closeness centrality is based on the number of hops it takes for a node to traverse the entire network, and it indicates a person's access to resources.

These centrality measures enabled Randle to view his organization from varying perspectives and identify important traits of both people and departments.