Bridges in Learning

Why bridging people, not just building courses, can lead to more effective training

Jose Villalta

3 min read

A bridge over a body of water at night
A bridge over a body of water at night

Learning doesn’t just spread through content; it spreads through people. While L&D metrics often prioritize tracking course completions and survey scores, much of what actually changes behavior happens through informal networks. This could be a tip shared over lunch, a new practice picked up from a peer, or a workflow passed between teams. Additionally, the people behind these informal networks are just as critical as the process itself. These folks quietly shape how new ideas travel by connecting teams, translating concepts, and modeling behaviors that others adopt without ever stepping into a formal training class. In a high school context, a study of 56 schools found that when highly visible students (those most observed by their peers) were encouraged to take a public stance against conflict, schools saw a 30% reduction in disciplinary incidents. These “social referents” turned out to be a good way to strategically influence a network.

But is there a way to identify these well-positioned people? One method is betweenness centrality. In network science, betweenness centrality is a concept that identifies people who serve as bridges in a social network. These aren’t necessarily the most senior, visible, or connected employees, but the ones who sit at the junction between different groups. In other words, they act as informal players who connect silos.

Betweenness centrality explained

Imagine Alex, a retail operator who works closely with both the store leadership team and the supply chain team. If these two groups rarely interact, Alex, being a bridge between the two, could help share updates, resolve issues, or align goals. Alex has high betweenness centrality, not because of a title, but because of her position in that network. If Alex also works closely with other teams that rarely interact, her betweenness centrality increases. This high betweenness centrality means she’s structurally positioned to disseminate information more easily between disconnected teams.

In a 2021 study of over 12,000 creative teams in the board game industry, researchers found that teams with higher betweenness scores (those that bridged otherwise disconnected parts of the network) were more likely to produce award-winning games and receive higher public ratings. These teams had faster access to diverse, non-redundant ideas flowing through the industry. The study showed that teams in bridging positions weren’t just more visible, they were better able to access and apply ideas from different parts of the industry, which boosted their creative output.

Betweenness centrality in an organization

Finding people like Alex can be a challenge. To start, you’d need a way of knowing who she interacts with. Proxies like emails, chat logs, or even meeting invites could be a good place to start. Then, you'd need network science tools that map out those connections as a way to identify who talks to whom, and how often. The data would also need to be cleaned, transformed into a network of nodes and edges, and analyzed using algorithms that calculate metrics like betweenness centrality. More importantly, each step, from choosing a time window to defining what counts as a “connection,” can significantly shape the insights you get. Given the effort, it's no surprise that most organizations don’t have that kind of data readily available.

Luckily, a more accessible starting point exists: running a simple survey. To start, poll a group of employees from various parts of the organization and try asking them, “Who do you regularly go to for advice?" or "Who regularly helps you with work problems?” The people who show up frequently, especially across different teams, are likely informal brokers of knowledge. The insights gained from this exercise can be a powerful first step in understanding how learning can flow across the organization.

Once you've identified these informal brokers, equip them. These may be the people best positioned to spread ideas organically, so consider prioritizing them when rolling out new training, frameworks, or cultural initiatives. Bring them in early, give them context, and ensure they understand the “why” behind what you're introducing. Their buy-in can accelerate diffusion, especially when trust and influence already exist. Even a short, focused session with this group may have more downstream impact than a large-scale training push, simply because of where they sit in the social fabric of your organization.

Limits

To be clear, structural position and high betweenness centrality don’t guarantee behavioral influence. Some brokers may be uninterested in sharing knowledge or lack trust among peers. There's also the question of fairness. Only training some employees means not training others. Additionally, recent research suggests better methods of diffusion. A study by Saqr & López-Pernas found that diffusion-based centralities (like coreness and diffusion degree) were more effective than traditional metrics like betweenness at predicting learning spread. These metrics captured not just connections, but how well-positioned someone was to actually transmit and reinforce knowledge through collaborative interactions. Still, these insights give learning leaders another approach to training. By identifying the right, not just the next ones on a roster, organizations can build smarter, more scalable strategies for spreading knowledge.

Final Thoughts

With limited time and resources, learning leaders need ways to amplify impact. While not perfect, betweenness centrality and network science more broadly offer a practical way to do that. It helps shift the focus from what to teach to who to teach and when. By complementing content with strategy, we’re not just informing, we’re empowering.