posted 10 Oct 2005 in Volume 9 Issue 2
Social-network analysis Part IV
Practical examples of how different leadership and KM practices improve overall connectivity in an organisation. By Patti Anklam
“So now what?” is the big question following a network analysis. You’ve done the planning, got the stakeholders engaged, decided what questions to ask, completed the survey and the analysis and you have some interesting results to review. In and of itself, you’ve just completed a great deal of work, but in the overall scheme of things you’ve just started a process that is intended to move an organisation – or set of organisations – from a current state of connectedness to a state of improved connectivity and awareness of the importance of being connected. This fourth and final part of this masterclass series provides a framework for thinking about the role of social-network analysis (SNA) in the context of an organisational change effort designed to improve collaboration, enhance networked thinking, or solve specific business challenges.
First results, first response
The first people who review the results of an SNA are the client sponsors – usually the senior managers in a group. Those who were sceptical at the outset of the SNA will first look to see if the results really match their own perceptions of the interactions of people who are in their departments or sections. Invariably, what they will see is that the SNA has generated data to match their perceptions. One manager recently commented to me on an individual who was in a good position with respect to his peers in an organisational network analysis (ONA); this person had been promoted just at the time the survey went out. The manager said, “This confirms that I made a good decision in this promotion.”
In this particular case, the SNA was intended to give the entire organisation a baseline of the current level of collaboration across divisions. Because the data collected included the demographics of the organisation, it was possible to draw charts and show metrics for each of the divisions individually. At both organisational and departmental levels, the SNA data can provide rich insights – and lead to critical questions.
Conversations for change
The ONA consultant’s job is to provide insights into the themes that emerge from the data, and recommendations based on those insights. In Part 3, I reviewed some of the techniques for illuminating charts and displaying metrics based on demographics that can prompt probing questions. In a preliminary review of results to a client, I typically have 40-50 charts that are organised based on the themes that emerge. Some typical themes that recur are:
Lack of connection across groups based on a demographic attribute – the division or department, role, tenure, geographic location, level of hierarchy, and so on;
One or more people being very central to the organisation – usually senior managers who are very good communicators will show up on the charts as both giving and receiving information from a disproportionate number of others;
Results from specific questions that are particularly telling – for example, if the question, “How well do you know this person’s knowledge and skills” has a very high density, then a theme would be that people are not aware of each other’s skills;
A hierarchical pattern of advice seeking or looking for input from divisions – that would reveal an organisation that may be excessively bureaucratic, or just even more bureaucratic than it thought it was;
Figures 1 and 2 are examples of two of the types of SNA metric results that have the greatest impact and can lead to the most interesting conversations.
In the cross-boundary analysis example, practice managers for a global professional services organisation participated in an SNA. One of the questions they responded to was, “How effective do you think you could be if you communicated more with each person?” The density values of responses within each job level are shown in the matrix. (Density in SNA is the percentage of actual ties that exist out the total that are possible.) Here, we see that the density of response among the global practice managers was 41 per cent. Among the regional practice managers, the resulting density was 10 per cent. From the data, it appears that the global managers would like to have more communications among themselves than do the regional practice managers. Is action required? It depends, of course, on whether there are in fact opportunities for leaders across geographies to share best practices, swap ideas and learn faster about what works at the local level. If there is opportunity for sharing, then some form of communication is required – that probably begins by bringing people together, to meet, and to understand their common areas of interest and the potential for each to improve local performance by leveraging global counterparts.
A scatterplot enables you to look at the data from two questions. In this case, the survey asked about the frequency (number) of interactions that people had with each other. A separate question asked whether interactions with each person had significant value with respect to personal performance.
Here, PB (the group vice president) is shown to have many interactions with most of the people in the group, all of whom value those interactions. But perhaps PB is taking too much on himself? Perhaps he should look at those (highlighted) who are in the centre of the chart as people who may be ready to take on more interactions. An important response in this instance would be for this manager to ensure that job development and performance plans for high-potential performers include more opportunities to interact with people across the organisation.
A network map showing connections among people in a large and distributed group can look very positive. That’s why, when we do an analysis, we look at many different views to see which ones are the most compelling. In organisations that have worked hard at or prided themselves on being very well connected, it’s often useful to show the “0” responses to a question. For example, if the question is, “How frequently do you receive information from this person that you need to do your job?” and survey respondents are asked to place “0” if they don’t know a person, the map that shows only the “I don’t know this person” responses can prompt questions about where the integration needs to occur. Note that in this map (figure 3), those who are the least known by others appear in the centre, whereas those most well-known across the organisation appear on the periphery.
Facilitating the group response
During the first conversations, it’s important to validate the themes with the SNA sponsors and to notice which charts they ask the most questions about. A very analytic client may ask for you to create additional views; these requests will also indicate where the most meaningful interventions might be for this organisation. You may need to make some iterations, but it is important to hone the chart set down to a small enough set that is agreed to be meaningful and can be reviewed in a large group in order to move toward effective action.
Review of the results of an SNA can be an energising experience for an organisation – as long as the managers of the group maintain a commitment to leverage the results for improvements that will benefit the survey participants. In a client setting, it is best to work with an experienced organisational development (OD) practitioner who is familiar with the organisation both to finalise what is presented and what questions will be used in small-group breakout sessions.
These questions should focus on positive, action-oriented questions, for example:
Which departments could benefit from more cross-group information or expertise exchanges? How might they go about doing that?
What is the cost to the organisation of people not knowing what each other knows? What could we do to get to know each other better?
What mechanism should be put in place to ensure that people who are new to the organisation are introduced more quickly?
Are there decisions that don’t need top management approval?
It is also very good to include some activity that will enable people to focus on their personal networks and their individual interactions with others. For example, if a survey has asked the question, “How do interactions with this person affect your energy level (Energising/Neutral/ De-energising)”, then you could facilitate an exercise in which people are asked to list the key people that they go to for problem solving and then to evaluate (for themselves) what effect they are having on each person’s workload or energy level, and whether they are in fact talking to the right person.
Remember that it is never a goal to connect everyone to everyone else, but to ensure that the connections that ought to exist do exist; that critical connections are not being overloaded at the expense of enabling more cross-group connections; and that both individuals and the organisation are working optimally within the context of the network.
In their work, Rob Cross and Andrew Parker1 identified three main categories of response to an SNA. When you work with an organisation, it’s important to keep these distinctions in mind. If the presentation of the SNA project is to include your recommendations for actions, then it will help to think through these categories and the kinds of questions that lead to making decisions about change (Table 1).
Activities in some or all of these areas will result from an ONA project. The ONA project itself may be a one-time intervention to work on a specific business problem, or may be the starting point for establishing a new set of business-performance indicators. A key benefit of using SNA is that it enables managers to target specific patterns or metrics that need attention. I’ve not heard of an SNA resulting in any large scale organisation changes (re-organisations or re-structuring). What do occur are subtle but important shifts in personnel, infrastructure improvements, or communications and work practice improvements.
Conclusion: A framework for net work
This series of articles has presented views, examples, and tools for:
Getting the context right and framing a business problem for which SNA is the appropriate tool;
Designing and managing a network analysis project;
Getting to the “good questions” – reviewing and interpreting the data from an analysis to present data meaningfully;
Focusing results on both individual and organisational performance and metrics.
It is important, of course, to conclude with a reminder that SNA in and of itself is only an assessment tool and should be used in the wider context of a business or organisational goal. SNA can tell you what connections exist and it can tell you a lot about those connections. The conversations generated by the SNA can lead to an understanding of the relationships among the organisational mission, or context; the degree of relatedness among people and the extent to which social capital creates and sustains a motivated workforce; the potential for enhancing collaboration and communication through technology; the introduction of work practices that integrate connection and communication into daily activity; and the impact of organisational boundaries, roles and responsibilities on the nature of work. These five elements (as shown in figure 4) need to be in alignment, and checked often for misalignment. SNA is an excellent tool to uncover the human dynamics that are at play in each of these systemic elements of a well-functioning organisation. n
1. See The Hidden Power of Social Networks, Rob Cross and Andrew Parker, Harvard Business School