Ego Network Analysis — The key to enhancing Employee Social Capital

Organizational Network Analysis (ONA) has become a key tool for leaders to visualize and analyze the Social Graph of a firm. Some of the specific advantages of doing ONA are >> 1) Understanding key communities in a firm 2) Understanding who are the “central” employees in a firm. 3) Identifying employees who are “isolated” and hence need to engage more effectively and many more. Here is a primer on ONA.

However, while ONA is good for understanding the network as a whole — there are multiple advantages of studying networks surrounding an individual. These are called “Ego Networks”. In other words, John’s Ego network would be a graph showing his friends, his friend’s friends, his friend’s friend’s friends etc. The key point is, this network has John at the centre and is primarily meant to understand the support system he has at work (or even outside work). Here is a primer on Ego Networks. The central individual in an Ego Network is called the “Ego” and connections are called “Alters”.

Why study an Ego Network?

There are primarily two reasons for studying Ego Networks (as highlighted in this insightful YouTube video from James Cook @ The University of Main at Augusta)

Reason 1) Individuals are Interesting! — In today’s hyper-connected workspace, organizations are constantly trying to identify how to enhance the Social Capital of their employees by identifying the issues (or opportunities) in front of employees on account of the constraints ( or benefits) of the networks around them. There are multiple direct applications of Ego Network Analysis

1. Leadership Development — Identify how well leaders are positioned to leverage their networks as support mechanisms, information pathways, and innovation channels.
2. High Potential Programs — Use the Ego Network of potential HiPots to quantify how effectively they are cultivating their relationships and developing themselves as leaders.
3. Social Capital — In general, the Ego Network of every employee can provide a wealth of information around areas of improvement like quality/quantity of connections, diversity of relationships, and level of support provided by colleagues.

Reason 2) Complete Networks get very large, very quickly! — If we are to do an ONA analysis of a midsize organization of 5000 people, assuming each person has 15 immediate links (i.e. First Degree connections) — that’s a network of minimum of 75,000 connections! And this is without considering Second / Third Degree connections (i.e. friends of friends). Given the size of data involved, there is no way to do this analysis without the use of tools like OrgLens (www.orglens.com). One way to start the analysis is to select a representative sample of individuals and analyze their Ego Networks.

Types of Ego Networks

There are primarily 3 types of Ego Networks:

Level 1.0 — These are Ego Networks where we only track the immediate connections of the individual.

Level 1 Ego Network

Level 1.5 — These are Ego Networks where we track the immediate connections and connections between those immediate connections.

Level 1.5 Network

Level 2.0 — These are Ego Networks where we track immediate connections, connections between them and 2nd degree connections (i.e. friends of friends)

Level 2 Network

Why stop at Level 2.0?

Beyond 2 connections there is a significant dip in the quality of relationships and people are rarely aware of important details. This has been confirmed in a research paper by Noah E Friedkin in which he emphasizes that there is a “Horizon of Observability” beyond 2 connections where the quality of data is just 10% (and dipping for every additional step).

Key Concepts

If you’re interested in studying this topic further, here are a few key concepts that you should be aware of :

1. Geodesic (for each node pair) — Shortest distance between two nodes.
2. Network Diameter — This is the “size” of the network and is the largest Geodesic.
3. Homophily — Is there a tendency to limit one’s network to people with similar demographic features (eg: Same Gender / Profession / Age Group etc). This is particularly important in organizations spread across multiple geographies or cultures where diversity in one’s network is critical for operational efficiency.
4. Ego Network Diversity — Review the Ego’s contacts to quantify how diverse are the characteristics of the contacts. A popular framework used for this is Blau’s Heterogeneity Index.
5. Clustering Coefficient — The tendency of an ego’s friends to be connected with his other friends is called Clustering and the metric that quantifies this is called Clustering Coefficient (CC). A low CC puts the Ego in a advantageous position because he is the only bridge between a large group of friends but also has a disadvantage of making the Ego a bottleneck wherever the entire group needs to collaborate. A low CC increases the possibility of Structural Holes — a concept pioneered by Ronald Burt.
6. Brokerage — This highlights the nature of the Ego’s directional relationship between 2 Alters and is used to classify the Ego as Coordinator, Consultant, Gatekeeper, Representative or Liaison.

To summarize — Ego Networks are key to understand how individuals are constrained or enabled by the network around them. Also, if we want to understand the variation in behaviour between individuals (from a network perspective) then conducting Ego Network Analysis is critical.