Capitalizing on communication

Valdis Krebs is a pioneering social network analyst whose software product, InFlow, maps relationships in organizations and helps teams improve their effectiveness. Gerry Falkowski recently retired from IBM Consulting and now teaches at the University of Minnesota. Krebs and Falkowski have worked together for over a decade, using social network analysis to improve communication in enterprises. InfoWorld Test Center Lead Analyst Jon Udell spoke with them about their insights into enterprise social networking.

IW: What does it mean to “do social network analysis” inside an organization?

VK: There was one consultant … [we worked with] who specialized in helping newly placed executives get up to speed. She would work with execs who had just been hired for a high position, C-level or VP, in an organization where they had no experience and no network. To hit the ground running, they needed to understand the organization. We would use InFlow to map it out. These people know this; these people know that; there’s a cluster. And we’d put together a strategic plan.

GK: My last project was with IBM’s On Demand strategy group, a leadership team of 23 men and women who report to the CEO. I had them stand in a circle, I took some black yarn, and I connected the strong links. There were only six. Then I gave everybody three pieces of yarn and said, “Who do you need to connect with next week?” Picture these people exchanging ends of yarn with one another to form a web. Then I got one of the leaders to stand on a chair and look down at that web. “That’s what your team needs to look like,” I said.

VK: Of course these are people, not Cisco routers. If John and Mary just got divorced, they’re probably not the best choice to connect sales and marketing.

IW: Many SNA tools tend to gather relationship data by e-mail surveillance, but your method is survey-based.

VK: That’s right. What Gerry and I do is explicit. We come in with a set of specific questions that will elicit the information we need. It’s not a fishing expedition. We’re looking to solve a particular management problem. Of course, if you have relationship data from contact management systems, you can feed that into the tool.

IW: These projects were about strengthening ties within groups. Where does Stanford University Sociology Professor Mark Granovetter’s “strength of weak ties” idea fit in?

VK: People have a wrong impression about what a weak tie is. It’s not just a casual acquaintance. A weak tie used to be a strong tie; there was trust and shared knowledge. If I go to a conference today and meet somebody new, people will say “That’s a weak tie.” I say no, it’s an acquaintance tie. But if I also run into Peter from Disney, who I used to work with, that’s a weak tie that can be reactivated. A lot of Granovetter’s research on weak ties was based on people who had known each other better before.

IW: Social network analysis can reveal that highly connected people are more valuable than the org chart or salary plan suggests. Is this becoming a factor?

VK: Yes. I did a project with an investment bank, and they took into account who was most valuable in getting a deal done, and factored that into the bonus. I’ve had execs inside and outside IBM saying, “If this data is true, then I’m not paying the people who bubbled up to the top what they’re worth.”

IW: Does it cut the other way, too?

VK: We wouldn’t take a job that we knew would lead to a resource action.

IW: Resource action?

VK: Layoff.