Information Silos’ Impact on User & Customer Experience

Information Silos are… wait for it… bad (I know, you’ve heard this one)

I’m not going to spend a lot of time explaining what information silos are, why they are detrimental to any growing engineering organization, or how we can prevent or dissolve them. These topics have already been discussed at length by people who are much better at explaining them than I am (I really like The Silo Effect by Gillian Tett, if you’d like to hear more on that). To quickly summarize, silos are teams, departments, or other organizational units in which information becomes trapped. Information doesn’t flow freely between silos, resulting in decreased worker morale, slashed productivity and efficiency, and stifled collaboration and innovation.

Many of us in software organizations are well aware of the concept of silos, their internal symptoms, and the dangers they pose. We talk about silos a lot (so much so that I almost decided to scrap this article out of fear of exhausting an already-tired subject), but when we talk about silos, we almost always limit the conversation to only include the impact they can have on internal organizational performance. I believe we need to expand the conversation to consider the impact that these inefficient internal structures have on customers and end-users. After all, we are user-centered, right?

To be truly user-centered, an organization must see every decision through the lens of its impact on the end-user or customer, no matter how small or how far in the background it may seem. For example, we make tech stack changes so that the end-user receives a more responsive, secure, or otherwise “better” product. We adopt services like Kubernetes so that we can more quickly scale and deliver value to users. These are changes that are seemingly invisible to those that are actually experiencing our products or services, but we must consider their impact on those experiences. If we can’t map a change all the way back to the end user then that change is not worth making, and if we can’t measure the impact of a problem on the end user then that problem is not worth dedicating resources to solve.

I’ve written all of that simply to say that, as truly user-centered organizations, if organizational silos are truly a problem worth considering then they must have an impact on the end-user. Recognizing those impacts are important for a couple of reasons:

1. Being able to spot the symptoms of internal silos that manifest themselves in an external UX or CX can help identify those silos as the root problem, preventing us from solving the wrong problem.
2. Realizing and acknowledging that organizational silos can result in a poor experience for the end-user, and can even prevent user-adoption, can inspire us to be more intentional about preventing those silos from forming in the first place.

Silos Create an Inconsistent, Unclear, and Disjointed User/Customer Experience

Conway’s Law and Information Silos

Conway’s law is an adage originally authored by computer programmer Melvin Conway, who suggested that the design of any system will mirror the organization’s internal communication structure. For example, think of an organization that decides to build a product and to split that product’s functionality into four sections. This organization splits personnel into four teams, each focused on one of the four sections of functionality. Naturally, the structure of the user-facing product will be based around these four (potentially arbitrary) sections of functionality, regardless of the needs or mental models of the end users.

Any organization that designs a system will produce a design whose structure is a copy of the organization’s communication structure. — Melvin Conway

We often cite Conway’s law in a humorous or tongue-in-cheek way, but several studies (like this one conducted by a research team at Harvard Business School) actually support its premise. If Conway’s law is, in fact, a “law” then it stands to reason that a heavily-siloed organization, in which information does not flow freely from team to team, will produce systems through which a user can not freely and easily navigate.

For a short time in 2018 and then again from late 2019–2021, I worked at Kessel Run, a software development unit within the US Air Force. During my time at KR, I was able to witness Conway’s law in action in a couple of different situations. While there, I saw the organization grow from a small startup within the DoD into an organization with over 1000 personnel assigned to it. I watched the organization accomplish some incredible things, but as is the case with any org that scales that rapidly, KR also experienced some growing pains that we were able to witness and learn from. One of those growing pains was undoubtedly learning to share information across a larger and more disconnected organization. This pain eventually manifested to the end user in a couple of different ways.

Disconnected Teams Result in Disconnected UX

I first encountered the effect of information silos on a user’s experience when I went on a research trip to an aircraft maintenance unit to witness users experiencing our product in a real world context. This product was split among ~10 development teams, all focused on delivering a small piece of functionality. One product team built a personnel management feature set, another product team built a scheduling feature set, etc. Another team then built a dashboard to connect each of these individual feature sets. Users could use this dashboard to navigate between feature sets. Despite being split into pieces of functionality, the organization intended for this to be one product from a user perspective. Being brand new to the organization, I was not on one of these product teams during our research trip, but was there to observe users from a high-level, portfolio-wide perspective.

What we observed on this research trip was interesting: while each of the different individual feature sets tested well and received generally favorable reviews from users, when we asked those same users whether they were ready to adopt the product as a whole the results were mixed. While each of the individual features sets were great, users expressed frustration around having to use several different apps to complete their jobs rather than the one app that they use now.

But wait… this was just one app, right? Not from a user’s perspective. To navigate to a different feature set, the user had to go all the way back to the dashboard. There was no one-step navigation from one feature to another, causing each piece of functionality to feel very disconnected. Each feature set also differed slightly in how it looked and felt. One feature set was using a table for sets of information, while another app was using cards. One was using a light theme, while one was using a dark theme. One had a “share” or “export” function while another didn’t. All of these small differences added up to present users with slightly different experiences for each feature set. These slight differences meant that users had to learn how to navigate through up to 10 different interfaces instead of just one. It also meant that every time they needed to use a different feature, there was additional context switching needed in order to navigate through this new feature. The experience was anything but seamless, imposing a much heavier cognitive load on our users than was necessary. This led to their hesitancy to adopt our product.

The problem had nothing to do with the usefulness or usability of each of the individual features. The root problem was the organization’s internal structure. Each product team was so hyper-focused on delivering value for their specific feature set (which, once again, they did successfully) that no one spent much time communicating with other teams about how it all fit together. Each team was a silo for information about the implementation of their own feature set. Teams knew what others were working on, but specific details weren’t really shared, common goals and conventions weren’t established, and the cohesive experience was never considered. This lack of internal cohesion and collaboration led to a very confusing and disconnected end user experience.

Thankfully, upon hearing these results, the organization made the changes necessary to make the user experience more consistent and cohesive. The org encouraged cross-team collaboration through organizational rituals, like demos and critiques. Leadership placed a noticeably heavier focus on clearly communicating higher-level organizational goals. They also regularly shifted personnel between teams to ensure that context was shared as widely as possible. Designers began to adhere to a common design language with a portfolio designer ensuring consistency across the entire product.

I was not part of the organization long enough to see the final results of these changes, but this provides a great example of an organization that was able to see the symptoms of internal silos present in the experience of the end user and address the root of the problem.

Lack of Internal Clarity Results in Lack of Customer Clarity

Information silos can also cause customer confusion by causing an organization to present an unclear idea of how to interact with its products or services. Another role I had at KR was as a service designer on a platform team. In this role, I focused more on the customer experience of an invisible service (in this case, the platform) rather than the user experience of a tangible, visible product. Still, the effects of siloed information and a lack of context sharing could be seen in the experience of platform customers.

One of the projects my team took on during this time included mapping the customer’s experience of getting support from the platform and recommending improvements that would help to improve that experience. Through our research, we found a common theme: customers really didn’t understand the organizational structure of the platform. They rarely knew who to go to for each type of problem or request. They would often submit a request or ask a question in a channel that felt like the right place for the question, hoping that they would get an answer. Sometimes they got their answer, sometimes they were redirected (and redirected again), and other times their request was dropped entirely. To them, submitting a request was like throwing it over a wall without knowing what was on the other side and hoping an answer came back. Because of the lack of clarity into this system, customer development teams often had their progress slowed because they were waiting for platform support that may or may not come.

To start finding ways to make this experience better for customers, we started interviewing the teams that made up the platform. We found that many of them actually didn’t understand the platform’s organization any better than our customers. It was hard to know which team was responsible for which issues. This was what led to customer questions getting dropped or redirected multiple times. Very few people in the organization understood the structure and division of responsibilities of the org outside of their teams. That prompted us to ask an illuminating question:

How could we expect a customer to understand our organization if our organization didn’t understand itself?

Once again, each team’s well-intentioned hyper-focus on delivering value through their own service offering led to a lack of understanding of the organization as a whole. A lack of collaboration and the presence of constant change had resulted in a collection of teams that didn’t understand how they fit together as a whole. This organization was also able to implement some solutions to address these issues, including standing up a team whose members would consist of representatives from all of the teams within the platform. This team not only provided customers with a central location to seek help, but also encouraged collaboration and context sharing by pulling members from different teams together to solve customer problems. By reducing silos and enhancing internal understanding of the platform, the organization was able to enhance customers’ understanding of how to interact with the platform.

Ok, I’m done

By this point, I hope that I’ve convinced you that information silos are not something that only harms an organizations efficiency and productivity. The detrimental impact of information silos can reach all the way to the end-user, causing bad experiences with our products. My hope is that, by recognizing the potential of silos to harm user experiences, we are better able to recognize silos as the root cause in those scenarios and to implement the appropriate solutions.