How do you know if silos exist in your company?

Silos exist in most larger groups. There is no way denying that. Sometimes these are beneficial, but most often than not, they hinder innovation, result in duplicated processes and slow down decision-making.

There are some signs that silos exist in your team:

* The same processes are implemented in different teams in the companies.

* The solution developed by team A is not adopted by team B.

* Limited creative potential of teams.

* Stuff takes ages to get done.

* People don’t know details about their projects.

While these examples could be a sign of silos, they could also exist for other reasons. Take the example of the limited creative potential of teams. Of course, groupthink can make teams less creative, but so is a manager who acts like a know-it-all, employee rewards that discourage risk-taking, or an organizational culture that shoots down ideas.

What data do you need to measure silos?

If you really want to know if there are silos in your team, you need to collect data about it. Some organizational culture surveys include questions about teamwork and how connected employees feel to those from other departments, divisions, or teams.

But these do not capture the essence of silos. Imagine your company has two teams, A and B. You work in team A. It’s a great team. You all get along very well. But you need input from team B to do your tasks. Now, team B, well, it’s “just” team B. It’s “those guys”. You get the idea? You might not think high of team B, but through coincidence, you get to know a person on that team. You’re chatting via Slack/Twist and have connected over a common interest. You end up regularly talking with each other.

How would you answer the question “Do you feel connected with team B?” You could answer “No”, that answer reflects how you feel about the majority of people in team B. But it isn’t a 100% accurate representation of your feelings. You are very well connected to one team member. You could calculate an average, but can you average out feelings across people?

If you really want accurate data about how connected you feel to team B, you need to answer this for every team member. This will give you a network picture, a clear map about how you feel about team B. The picture below is exactly such a map, a social network figure. You can see that most people from team A (red) do not interact with those from team B (blue). There is one link from team A to team B. This is you and your friend. Imagine what will happen if you leave your current role or your friend in team B leaves? You end with two perfect isolated teams. Of course, thanks to organizational processes, they will still interact, but only the bare minimum.

What metrics do you need to measure silos?

The example above highlights that if you want to know how connected or disconnected your company is, social network analysis is the best and most accurate way forward. Of course, the drawback is that collecting that type of data is more cumbersome for participants or, if electronic communication such as emails or slack is used, requires greater data storage.

Silos exist when members of different groups do not talk with each other or where communication is infrequent. There are different ways to measure if silos exist. You can take the characteristics of team members into account and check if, for example, gender, location, expertise, influences with whom someone talks to.

Rabbit hole into research on communication

I mentioned gender and location as two characteristics that could explain with whom someone communicates. There are two theories that try to make sense why gender or location play a role with whom we talk.

Homophily theory: Generally, people prefer to talk to those who are similar to them. This could be gender, ethnic background, functional expertise, preferred sport.

Proximity theory: Generally, people talk more with those who are near each other. You are more likely to talk with the person who is sitting next to you, than with someone who is further away. That’s why companies spread over several floors or buildings automatically create silos.

Based on these two theories, if you are female, are you more likely to talk with your male co-worker who sits next to you rather than with your female co-worker who sits in a different office?

Next to individual characteristics, other aspects can also be used to measure if silos exist in your company and what their impact is. This is one way to measure silos. Another way is to look at the communication patterns, regardless of personal factors. This gives you a bottom-up view of communication patterns free of any influences.

What to do with silos?

The big question is what you should be doing if your company has silos. Depending on your situation, you might not want to do anything. Maybe these silos are beneficial for your operations? Or at least the benefit of reducing the silos is lower than the costs you will have to incur working on diminished them.

If you want to reduce silos, you need to give your employees a reason to interact with each other. The interaction will - most often - not happen automatically. Employees need to learn to trust each other, share information, or ask each other questions. Building trust requires time. You need to put them in a situation where they can experience each other's knowledge, get familiar with each other and develop trust. But, as said before, you need to give them a reason to interact and reward that interaction.

How to break silos?

* Create problem-solving teams with members from different silos. The problem should be relevant for several groups, and hence a company-wide solution is necessary.

* Create shadowing or mentoring programs that pair of people from different silos.

* Random coffee breaks where members are from different silos.

* Ask people from different silos to organize company retreats or other events (birthday parties, celebrating successful projects, or employees).

Your next step

You need to decide how you feel about communication and collaboration in your company: Is it fragmented or cohesive? If it is fragmented, how much does it impact your company’s performance? Is the impact positive or negative? Until this moment, you are only thinking about the problem and using anecdotes and feelings to build a case around it. This is the first step, but you will need data to measure the extent of the problem. Without measuring it by collecting data through social network analysis, your solutions will not be fully adapted to your specific context.