Employee Network Analysis, a Powerful Tool For CEOs & HR

You will surely agree with me that workplace relationships are important. For many organizations, this is even the key thing they seek to support and develop in their corporate cultures. I know progressive HRists who went a step further before the pandemic during recruitment. They organized several lunch meetings with the potential candidate and his future colleagues. After such informal meetings, future colleagues were able to estimate very well how well they would establish an employment relationship with this candidate in the future.

The current home-office era has hit relations even harder. Accidental meetings in the company’s kitchens have completely disappeared. On the contrary, we call from our home kitchens, but the conversations are not so random and relaxed at all. We only see the head of man. It’s much harder than live to zoom each other. And just jump on a piece of speech even harder when you can’t invite a colleague for a beer or a baked cake at home.

Many relationships are still catching up with what they have saved in the past. This is probably best described by Steven Covey’s concept of 7 habits of highly effective people, the so-called Emotional Bank Account. Steven Covey explains that when people have a relationship, it behaves like a bank savings account. When you are kind to the other person, you show him understanding, support, or you are interested in him, then every such act is like a deposit in this emotional account. However, we do not always manage to do only the good and nice things. Sometimes our vices show up, sometimes we have little time and sometimes we just don’t notice each other’s needs and then we make withdrawals from this emotional account. So it is important that this account is in full. Many built strong relationships and saved on such an imaginary account. Today, they draw from it and those stronger working friendships can continue at a distance. The weaker ones disappear.

But how to properly analyze the current situation of relationships in your company? Standard surveys measure the corporate climate or work atmosphere as a whole. From the results of the CEO he sees that whether the relations are OK on average or not. And if he wants to see more detail, then the results are evaluated e.g. by individual departments, which again only provide information on the average value of the group.

However, the hierarchy of managers and departments does not show the true dynamics of the relationship. On the contrary, it often indirectly introduces you to think hierarchically on the subject. We automatically assume that a manager has influence over his team and has authority appropriate to his managerial level. However, this is rarely the case. This is because relationships behave more than the network in the following image.

The organizational structure does not reflect the real relationships in the company.

Do you know how to map, get to know and understand relationships in the workplace? Do you know how to find out who your key people in the company are and with whom they have the strongest relationship? Which of your experts is at risk and may resign next month? How to ensure that important management information reaches all employees through informal paths? Or who is the dangerous employee who is most hindering your efforts for your corporate culture or ongoing change management? And from whom does the toxicity in your company spread the fastest?

If you have the same or similar questions, the answer may be network analysis. You are on the right article. Read on.


Employee Social Network Analysis

Network analysis is not a modern novelty at all, but it is increasingly used and the age of the Internet has brought a number of commercially successful applications in the mathematical field of graph theory. You use the well-known commercialized success of network analysis unknowingly on a daily basis. Google search. Its initial success is due to the website analysis algorithm (PageRank) developed by Google founders Larry Page and Sergey Brin at Stanford University in 1996. And since the patent for the algorithm was created at the university, so that Google secures exclusive usage rights, so he gave his university an exchange of shares of his company. In 2005, Stanford sold these shares for $ 336 million. Since 2019, patents have expired and PageRank has found application in many other areas, such as measuring the influence of scientists on the basis of mutual citations.

Pictured: Larry Page and Sergey Brin in 2003

However, network analysis also uses many other algorithms and procedures in the field of traffic design, measuring the interconnectedness of economies, modeling the spread of viruses, or the latest advances in gene research, … … and thus in the field of knowledge of workplace relationships. In our application to the work environment, we will refer to people in a network graph as balls (nodes; nodes) and their relationships between them by a line (links, edges; links).

So the question is, what can network analysis actually bring to your business?

Really a lot. And I’ll try to explain the potential benefits with a simplified example. For simplicity, I have chosen a single question for the model example, but no network analysis should be based on a single question. In a model example, I will show you 5 new perspectives for CEOs and HR that no standard survey of satisfaction or engagement can bring you.

One question. Millions of possible combinations.

The basis of the informal influence of a manager is trust. If the manager does not have this trust, then instead of the so-called Soft tactics influencing your colleagues must use the harder ones. And trust cannot be built from day to day and at the same time it is very fragile. Therefore, it is one of the rarest essences of influence.

In our model example, we will want to map relationships, but we will want to know only the good ones. Therefore, we ask all employees the same and especially simple question: “Which people in the company do you trust the most?”. From the answers of the respondents, we begin to form relationships that can be one-sided or mutual.

Martin trusts Peter (1), Peter trusts Martin (2), Martin and Peter trust each other (3).

Of course, there may be a situation in your company where 2 people do not mark each other (4). In our analysis, we will consider this situation to be a “cold” relationship or a relationship that is even negative, and these people may even show distrust of each other. However, in order to simplify the implementation of network analysis, it is appropriate in this example to focus only on positive relationships.

Naturally, we must extremely simplify our view of the company and the people in it. Otherwise, no one would know. The organizational structure is such the easiest example of simplification. We put people in departments. We assume that team leaders have an impact on their people. We expect one department to deliver outputs to another. We communicate to the inside of the department through the team of the leader of this department, etc. However, in network analysis, we have the opportunity to use mathematical procedures that playfully cope with complex calculations. This will reveal completely new perspectives on the dynamics of relationships in our organization. Let’s take a look at them.

General view of the company

Real relationships in the organization

You will see who is close to whom and, conversely, which people are further away. In our model example, e.g. HR or CEO will see that the network in their organization consists of two units (components). Much of the company is connected together. But the small part on the right is torn off and there is no bond with the rest of the company. On the left we see a group of 12 people, who are connected to the rest of the company by only one person. What happens if this one person leaves the company? Are the people and teams that work together close enough?


A look at the communities in your organization

Let’s take this very simplified example of a company that has 7 employees and the trust between the people who are connected by a line is mutually positive.

Without us recalculating this network using mathematical algorithms, it is immediately clear from the picture that 1, 2 and 3 together form a better “game” with each other than with the rest of the organization. You can see the same “game” at the opposite end of 5, 6 and 7. Such games are called clusters or I prefer the word community. If employee 1 omits information in this community, then the probability that information will successfully reach employee 7 is significantly lower compared to whether 2 and 3 will know about it.

It’s actually something very similar to a phone game where you have to whisper the person next to you a secret link. Information always spreads faster and better in communities. On the contrary, if some information is to go to people too far away across the entire organization, it may be distorted, or it may evaporate completely (see image above).

The community can stick together and support the strategic activities of the entire company, or, conversely, it can hinder them. But most importantly, a community doesn’t have to be defined by a department or an organizational unit at all. Evaluating such a simple picture with 7 employees is relatively easy and visually. Thanks to mathematical procedures, however, it is not a problem to clearly identify communities across the organization, which can have thousands of employees. Subsequently, we will use different colors to visually distinguish the communities, and together with the assigned names, such a map of relationships and communities begins to give high added value. HR and the CEO himself, see in which communities the CEO or his managers are located. What are the distances between people and whether some communities are not too far from the rest of the organization.

Diversity, homophilia and inclusion

In the picture above you can see 4 communities: blue, green, purple and orange. At the same time, you will immediately notice how the individual communities are grouped together in color. Then it is necessary to warn whether there is no homophilia. Homophilia is a situation where people with the same characteristics form significantly more relationships with each other than with people with different characteristics. A generally socially widespread example of homophilia is when e.g. homophilia is created on the basis of race or nationality. However, in your businesses, homophilia can have much more subtle characteristics and the impact on your performance can be catastrophic. Let’s imagine e.g. homophilia on the basis of age when older colleagues do not let in young newcomers. Or they won’t let go of a talented and skilful woman in a male community. Or separation based on time in the company, education and political views. Of course I forgot something. It is also bad when people separate on the basis of nationality, sexual orientation, religion, etc. However, I intentionally use the word separation and not segregation, as it can be a natural and subconscious phenomenon when a crow sets a crow, seeks an equal. And separation can be natural by completely different circumstances, e.g. colleagues from Bratislava did not have the opportunity to meet with colleagues from Košice last year, and thus there was a greater separation of people in the company, based on location.

A separate article could be written about why diversity and inclusion are so important, I would just like to pack 3 nice statistics:

  1. Harvard Business Review: Companies with above-average diversity had 19% higher revenues from innovation.
  2. AmericanProgress: Inclusive companies have a better chance of gaining and dominating new markets.
  3. US Dept. of Labor – BLS: Companies with high inclusion are up to 120% more likely to achieve their financial goals.
Separated vs. Inclusive communities

Many companies have diversity strategies. They fight against various prejudices already when selecting employees. And it’s not only right but also good. Network analysis offers a concrete way to map whether relationships in the organization are successfully formed between people with different characteristics or only the company has managed to create separate groups under one common roof.

The example in the following image shows a part of a company in a smaller circle, where you can see a community that is cut off from the rest of the company and has only 2 relationships with the company. This community is separate. In a big circle, there are communities where inclusion works well – communities and employee relationships are intertwined. As a result, teamwork, creativity, innovation and communication work better. I think all companies would like to benefit from these benefits in their business.

Endangered individuals or entire teams

Endangered individuals and teams.

In each company there are individuals or smaller teams, which could be called the critical business infrastructure of the organization. Although the rule is that everyone is substitutable, the question also applies to how much you are willing to pay for it.

Studies show that the loss of a productive employee costs the company a hidden cost of 6-12 salaries until it is fully replaced by new employees. In more sophisticated businesses or highly competitive environments, the loss of an employee can mean not only a decrease in performance, but also the transfer of know-how to the competition, loss of key customer relationships or the inability to replace a person in the market, because in short the profile is extremely unique and unique.

Network analysis can also help you here. E.g. in the picture we can see that the community of four orange colleagues (right) is without a connection to the rest of the organization. Who are these people in our company? What would happen if this whole team was headed by our competition? However, they often do not have to be small groups at all, but they can be individuals who have only a very weak connection to other colleagues or do not even have it at all (eg a green ball in the middle up). Such people should be informed by their team leader, who can intensively intensify the work on building a relationship and the inclusion of the colleague in the wider team.

The power of influence

Let’s look again at a model example of a company with seven employees and an example with a game on the phone. What if in this example, employee 4 starts broadcasting information? What is the probability that the information will reach the employee 7? What is the probability that the information will reach all other colleagues?

The probability is higher than if the same information was sent by employee 1. In this case, we are talking about the impact. On a simple example, it is again evident that employee 4 has the most influence. The influence of individuals is determined by their position within the network. To put it simply, those who are on the fringes of the network usually have less influence than colleagues who are more in the middle. However, more sophisticated algorithms also allow for connections to other influential people in the organization. And so e.g. The CEO does not need to have a direct influence on all employees – it is enough if he has an influence on the most influential ones and thanks to them he gains a part of their influence for himself as well.

For a simple example and visualization, in our model example we will enlarge the balls according to the influence of the given employee within the organization:

Mapping and illustrating the impact of individuals.

This raises important questions about the complexity of your organization:

  • What influence does the CEO himself have?
  • Who are the key managers with the most influence?
  • Who are the important managers who should have influence but do not have it today?
  • What feedback or training do they need to be able to start actively strengthening their informal influence?
  • Who are the ambassadors in each community and how to use their influence to strengthen change management or communicate strategic initiatives?

More insights and what to look out for

Watch out for the imagination!

Network analysis, when combined with other data you have about employees, can bring you a lot of new inspirational insights and understanding. From your internal HR systems, you may see employees with high growth potential. If you notice that a person also has an informal influence, then you may have just found a great candidate for promotion. Or, conversely, do you know your problematic employees and want to understand the impact of saying goodbye to them within your organization? Network analysis hides many answers. Thanks to mathematical methods and visualization of relationships, you will get an overview that no other internal survey has given you before. Feel free to turn to experienced consultants who will help you combine the perspective of mathematical algorithms with the world of human relations and give you an unbiased picture of your organization.

An unbiased image means not only that they show you what you didn’t know, but that they also alert you when you see something you shouldn’t see in the charts.

Many leaders like to say that their people are the most valuable thing they have in the company. And relationships are the most valuable thing we humans can work on in life. Every successful leader must know how to work with this complexity. Network analysis is a tool that not only simplifies this work but also enriches it with views that cannot be created in any other way. If you are interested in the topic, do not hesitate to like, comment or contact me directly.

I look forward to interesting discussions.

Andrej Winter, Employee Experience Consulting