Full Circle Associates

(warning: my longest blog post ever)

I'm a member of CPSquare, a Community of Practice on Communities of Practice. (Yes, I love "meta.") We have an online asynchronous space along with an ongoing series of global teleconference calls on a range of topics.

Today Patty Anklam is sharing her research on social network analysis, with a particular interest in communities of practice, with our community. There are about 20 people on the call (with more who joined in progress) from many points on the globe.

Patty introduces herself as a researcher in deep learning mode. Jenny Ambrozek has contributed to her thinking. A lot of her thinking started when she joined CPSquare 2 years ago (the community just celebrated it's second birthday.) Testing the notion with people that networks are everywhere. We know that they matter. The more she looks at the analysis of networks, she is convinced the awareness of networks informs us as individual and organizations in the world and how we can leverage them better.

Networks are the substrate of livelihood where we live all the time. We have to take care of them, understand them. Learn about them. Key insights:

• we don't do anything alone anymore. I everything we do builds on the work of someone else. If we look at all the thought tendrils, there is a network behind that.
• Strong networks are correlated with health. People are healthier and happier in their work if they have good interaction with other people. Companies that know how to manage alliances are more flexible and adaptive.

Much of what she is talking about has been written up in a report that will be published by the ARK Group some time next month (March).

Networks matter and networks are everywhere. One of her forthcoming charts is a firstcoming cut at networks that are visible, invisible, formal and informal.

When she went freelance she joined some new networks. She saw distinctions. What are the different signs of networks? Teams are defined around processes, etc. The great work of Etienne and Bill around the elements of communities of practice, the role of boundaries around learners and practitioners. Very powerful. But also think and try not to fall into the trap of thinking everything is a network. A network is anything that can be described via nodes and ties.

So communities are networks. CoPs are special types of networks. A personal network we participate in all the time. Families, jobs, alumni groups, professional associations, informal social networks, job networks, support groups. Then there is this other emergent learning network she became interested in. It is not bounded. It has an identity -- people are interested in learning as a general topic area, but perhaps not tightly tied to their work practice. What is the model of this network? Being able to participate in such a network is an important piece of who she is and what she does.

It is a network in the Boston area, intentionally formed network of professionals, mostly consultants, who started from the standpoint that if we know each other well we can bring a diverse sent of viewpoints to solve problems. A consulting network with informal boundaries and governance. Were not able to sort through how to work with each other, but because they had evolved a complex set of valuable resources, they became a learning network. Many are interested in networks themselves. Over three years they have shared with and learned from each other. Not all in the same business, so diverse. Intentionally don't do social events outside of meetings -- not the intent. Modest governance, but not highly structured. This is her area of inquiry. Somewhat related to CPSquare and its practice groups.

Etienne asks “how would I know if I can belong to it or not?”

Patty replies: "You would have to pay some money and would have access to the website. There is some governance. You have to be recommended by an existing member. You can’t come to more than two meetings without saying ‘I want to join.’ When we participate, we bring guests to meetings, when any one of us is trying to solve a problem, we tap Gennova as a network. So we tap the networks of the members."

Pasha: "This sounds like a formal, intentional organization. So how is that different from a network. What is your inquiry?"

Patty is asking what are the attributes and characteristics of networks like this. She has another colleague/friend who holds a two day event around a topic in organizational change or learning. There is a network around that, you don’t have to be a member, but there is a name around it, people show up – it has the sense of an expansive group. Is this a phenomenon that is common? Are there characteristics of these networks that enable them to persist? How do you mange them? Should they persist? How can you both provide the power of a network and the place for people to learn.”

Susan: Have you read Fifth Generation Management by Charles Savage, how you’re defining a network. How do I know if I want to convene a network. How would I describe that to a group of people?

Patty: That’s the enquiry. What does it mean to create a network? The answer comes along multiple lines. What kind of governance? Infrastructure? Characteristics of members? Is it a network or close partnership? Blurring boundaries across individuals and groups and how we distinguish them.

In each network we bring a different identity, set of knowledge and skills at any one time. This notion of identities is becoming very important. Patty is working with people studying social physics of our online identities. Sometimes people will have different personas and names in different online networks, will want different levels of privacy and disclosure. So in addition to skills and knowledge identity, there is a technological set of issues around security and privacy of identities in different spaces. When you work in many networks you remain yourself, but people may see you differently in each one.

When you talk with someone you haven’t talked to in 10 years ago they may still be thinking of you as that person 10 years ago. For them, that is your identity until you update it.

Net Work – how do I manage my identities? How do I get past this magical 150 Dunbar constant of being able to manage many relationships at one time without cognitive overload. A key task is being able to know the status of my networks and relationships, where I’m spending too much or not enough time. Rob Cross at Univ of VA has been doing work looking at individuals looking at their networks within a domain and analyze the organizational effectiveness is impacted by where they spend their times in their networks. What relationships should be forged? Which should we spend less time with? Personal Net Work is important for people to understand the value of networks and how they can help us. Very new stuff. It is a new tool and we don’t quite know the results. Patty will post a URL.

Rob Cross makes a distinction between a network viewed from the outside and a view from a personal perspective from the inside. Of their personal network as compared to the network as an identity unto itself.

Networks from the outside end, social network analysis or organizational network analysis is Patty’s area of specialization. Diagnosing and intervening in organizations. Using a network view to look at knowledge and communications flows. Most are familiar with SNA at present. Looking at where people get their information to do their job can reveal network patterns. How people go to each other in an organization. She uses SNA to do this kind of diagnosis around different dimensions or questions, which help to understand the organizational flows. It offers insights into why things are happening they are and where things can be improved. It is not to take the initial survey map as fact, but to use it to probe with questions. Why is it that the person on the far left of the chart is only talking to a unique sub group?

From a managers perspective it helps to see who needs to be communicating with whom and how to improve that. Some of the data is collected through SN analysis software, pulling out data from existing software. Entopia pulls data from what content people access in a CM and then showing the proxy or inferred relationship of those working on the same documents. Can look at email, listservs, etc and detect pockets of expertise and connection. A diagnostic tool that gives an initial perspective.

The interesting part about understanding networks and how people are related in orgs, it leads to being able to manage the networks. Some say network management is an oxymoron because you can’t see the boundaries and it consists of constantly changing relationships. You can’t manage it like a predictable system. But it can help improve communication, knowledge flow, and organizational effectiveness. You can use the levers available from Knowledge Management (KM) to make a difference/change.

"Managed" is both the "inside out" and the "outside" in. She prefers the term leverage, because leadership is about how to leverage networks. Managing the context of networks and it is leadership. What is the business question and what is the network?

Etienne: It would be interesting to hear a snippet of a story. Are there ideal geometries? What would that mean?

Networks evolve. Geometries have different results. Valdis Krebs looked at an emergence of companies in the US Appalachian region where they started to collaborate. The first map shows very scattered clusters. At some point a person decides to step up to leadership around a topic or a program, and that person becomes the hub of a network with multiple spokes. That represents a map where one person is the key flow point. As the network evolves, more people become co-leaders or co-facilitators; you see more of the “small world” effects where you have small groups, clusters and topic areas. At some point you might move into the core/periphery structure – very familiar in CoPs.

Designing networks for optimal effectiveness – a paper is about to be published by Rob Cross et all about different types of networks for different purposes. They describe different work contexts and the amount of connectivity and leadership models as they appear across the different networks. You need to have a small amount of internal connectivity for fairly routine work. In something modular, like in professional services organization applying specific practices and processes and repeatable solutions, pulling together known sets of things would have a different network structure, with more external connectivity than an internally focused routinized structure. IN a highly customized system you have a lot of internal and external connectivity. Impacts organizational design thinking. There is a lot of research going on in this field.

Good leaders have always been about providing the context in which people can succeed. True about networks. The context of how a network is maintained, how there is enough diversity, appropriate mix of internal and external ties that make networks effective.

Etienne: What level of confidence do you have in the information you have received? When I have been surveyed, I was not confident of the responses I offered.

Patty notes there are various ways that people answer. You get generally sufficient validity. What she and others do is once you have the data, draw the maps, then you validate the data. Talk to people who are at the center, to those who understand the network. The survey is never the complete story and people will interpret things in different ways. The map provides questions, not answers.

Etienne: How would you do this for a client? Is there a practice? A platform?

Patty says there are both consultants and companies adopting this as a practice area. She spoke to a person at Haliburton who has adopted SNA as part of their team work formation process. They use SNA at the outset to identify existing ties and where there are ties that need strengthening.

Another of her colleague uses SNA in conjunction with project team evaluation. He is a team collaboration and project management expert. He looks at project tasks, timelines etc to look for areas of risk. Patty does the SNA to see if the appropriate communications lines exist and looks for rifts in the project and places to look at avoiding problems n complex, high risk projects. For an example when there is one person at the center of a network and a possible single point of failure.

Getting the survey questions right is very important. Valdis Krebs taught her to ask the questions in a very quantitative way. Get specific to numbers and time. It’s not so much that you are asking them about other people’s relationship, but your self-report of your perception of who they talk to frequently or not. In a group of 40 or more people the trend and overall result is accurate enough to ask reasonable doubt questions.

John: Do you find or believe that people in these different structured networks have heuristics for detecting the structure. Do people have intuition about if they are in a hub/spoke, core periphery structure?

Patty does not think that have perceptions at that granularity, but good leaders tend to know what the structure looks like. They effectively use the network structure, even implicitly if not explicitly.

In her professional services organization there was a concern that people in practice groups were using the same set of tools within their groups and not a lot of cross group communications going on. The SNA proved him correct. Although he encouraged people to connect, he had not had the evidence that would wake them up to change. The power of the SNA, and the presentation of the results is when it is presented to the group and the group involved does the analysis themselves. Rather than using SNA in a command and control mode. Noticing that someone is not connected and fire them. We don’t want to use it that way.

What does a SNA consultation look like?
Coming in to an organization, the first part is what is the presenting or business problem. It is a team assessment, as in a risk analysis. Could be around understanding competencies, which do you go to to get information. She really believes that people don’t call them because they have a problem and they think SNA will help them, but it is a good diagnostic tool to add to their other interventions and techniques.

Some people believe SNA can help with the “lost knowledge” problem as people retire. We don’t know the impact on our organizational competencies. Who knows what? That leads to questions of whom you go to when you need to know about a particular topic.

When the problem is communicating, collaborating or sharing information, the question around information can take a number of forms. Frequency of information, probe around accessibility – how easy is it to get access to other people in the organization. What question to ask well is well described in Rob Cross’ book. What are the questions you want to ask? Subsequently looking at responses she would work with the sponsor – a manager committed to the transparency of the survey, the results and to make a different with the results. Look for validation. She gave the example of a 7-cluster set of CoPs across 7 countries. In looking at their map, there are people that show up as potential bottlenecks or connectors. What are they actually doing? You can only find that out by going and talking to those people. SO the map indicates who to talk to, but it can’t describe what is happening. They may be experts. They may be gatekeepers.

In the SNA you can ask who you go to for expertise and with what frequency. Then you can assess knowledge and accessibility. Then you can decide as an intervention is this person a connector –passing information freely and helping it flow. Or is this person overloaded and doing this in addition to their job or is it their job. Should they get reward? Job restructured? How does their position in the network take away from normal activities or overload them. Those people who are central, overloaded and not acknowledge could be lost by the organization.

The next thing to look at is what are you going to do about this. Recalling that managing SNA is about understanding how to manage in complexity. It is not about a whole scale reorganization or command and control. What are small things that you can do to start to change the relationships. Maybe a F2F community event to nurture more connection across the countries and geographies. Maybe the recognize the central people were bottle necks, so some roles and responsibilities and decision making patters to increase the connectivity. Her role as a consultant, is to go through that process to help the organization – it is about self awareness and understanding tools and techniques – organization and KM—that are available.

When you see patterns, what are the ways to change them - -to increase connectivity, leverage expertise, etc. What Patty realized is that the interventions in networks aren’t starting or new. It is about judicious and intelligent selection of known interventions with a network perspective. Enabling CoPs, offering connection tools like IM or distributed collaboration tools can increase connectivity. If you need more diversity, or change how individuals work, you look for the things to leverage those activities.

In her report research she interviewed those using SNA in a KM context. She heard over an over that they had more F2F meetings, installed expertise locators, etc. Think about KM interventions in the networks.

Communities of Practice can be looked at as part of a diagnostic. Developing and supporting them is also a powerful intervention. HP Labs has been looking at email as a way of gauging the extension of expertise around particular topics. Not sure how much is using in practice. They are looking at ways to leverage understanding the pockets of expertise and where new things are happening.

Etienne: What are your heuristics to make sure the process is transparent and not some sort of “big brother.” What we’ve noticed in CoP interventions that people need to know it is about them finding their community and developing their ability to participate.

Patty says “communicate, communicate, communicate’ before during and after. The senior manager has to say this is important and that the results are for improvement but also insures individual needs for privacy are acknowledged. Most cases managers are comfortable making public that if anyone does not want their name on a chart, none of their names will be revealed. The manager has to be committed to that. It doesn’t happen frequently that someone does not want their name up. But it is all or no names. Give them comfort. The other thing that is important is never to do SNA in a downsizing environment. It will automatically generate behaviors where people won’t participate. It has to be used in an organizational context where the organization is healthy and wants to be healthier or where everyone is committed to improvement. You won’t get valid results when you introduce fear and it is unethical.

Social software, if you take the definition as any software that enables people to connect to other people – there are tons of stuff out there. Blogs, wikis, discussion boards. Think of CPSquare without WebCrossing. What would there be? Logging in after a hiatus, I got a note from Sus, then another. It is a real place. I love this.

Network and leadership – where is this going? CoPs are at the center. Your work in CPSquare has laid the groundwork for the work in networks. It is more wild and wooly in the network area, but organizations are looking for ways to leverage networks. Network organizational forms will never replace and supplant hierarchical forms, but help leaders know when a network form is better than a command and control form. She has started to work with Dave Snowden and the Cynefin Center and their description of new organizational forms.

It really is an interesting world out here. She is blessed with a great network herself locally in Boston. And

Nancy: Have you done any work with cross organizational or extra organizational networks? Ad hoc networks?

Patty: Ad hoc networks is a kind of emergent learning network. Why do people network. There is still a purpose there. Haven’t done a detailed look at that.

Nancy: What I see is that there is a key place for social networks (and SNA) in social movements & political change. In Armenia for example, communities have been in a command & control hierarchy during soviet area. Now they have to do more for themselves – organize and continue their evolution. A village needs to think for itself. It may want to start “thinking” with another village. This starts to think of changes: some of which may be perceived as disruptive and/or generative. They feel the buzzing but can’t hold it. I wonder about role of SNA for ad-hoc emergent networks.

Patty has not been involved in the social and political work of SNA. This would be a case where you might not want to do SNA because of the sensitivity of individuals, but a proxy analysis. Groups to group analysis and map from there. The field of SNA has broadened fast. Patty has focused on personal and organizational networks. There are people working in health care around immunity, SNA in different kinds of organizations, NPOs for example. SNA came out of ethnographic research and still a lot of people going strong on that.

John: What happens when you don’t have a sampling frame to represent the mapping space? You get the list of people to survey. IN the case Nancy is offering we’d have to traverse the nodes of the network to find the people.

Patty suggests that the survey leaves “blank spaces” to add other names. You build the map in an ad hoc way so people can ID others outside of the bounded network.

The group then thanks Patty for her wonderful presentation and the call ends!