Breaking Silos: A Network Blueprint for Reconnecting Organizations
Best Practices to Re-Connect the Disconnected
Most siloed organizations don’t decide to fragment. Rather, fragmentation emerges gradually as teams optimize for speed, expertise, and local success. Network analytics makes these invisible boundaries visible, showing where collaboration quietly stops and where targeted intervention, not blanket “collaboration initiatives,” can reconnect the system (Cross, Parise, & Weiss, 2007). In practice, the most useful lens is not the org chart, but the pattern of ties across groups: who exchanges information, who coordinates work, and who can translate across boundaries.
Below is a five-step, research-backed blueprint for breaking down silos in a way that is measurable, sustainable, and realistic.
1. Diagnose the silos
Before trying to fix collaboration, you need to see where it’s actually broken. Network analysis allows organizations to identify disconnected components, tightly-knit communities, and the specific boundary gaps where information and coordination fail to cross (Cross et al., 2007). A practical way to operationalize “siloing” is to look at the balance of internal vs. external ties between groups, often framed using variants of the E–I (external–internal) logic to quantify whether groups primarily connect inward or outward (Krackhardt & Stern, 1988).
This diagnostic step replaces intuition (“those teams never talk”) with evidence—pinpointing which silos matter most and where bridging would create the greatest value. In real terms, you’re looking for patterns such as high within-team density paired with low cross-team connectivity, heavy reliance on a few bridge people, and “structural holes” between groups where collaboration would create disproportionate value (Burt, 1992).
2. Pick the goal: discovery, transfer, or joint execution
Not all cross-silo connections serve the same purpose. Some bridges are meant to surface ideas and expertise (discovery), others to move complex knowledge reliably (transfer), and others to coordinate shared work (joint execution). Research consistently shows that different goals require different tie designs: weak ties are often effective for search and discovery because they connect otherwise separate pools of information (Granovetter, 1973), while more sustained interaction is often necessary to transfer complex knowledge (Hansen, 1999).
This distinction matters because organizations frequently overbuild the wrong bridge. For example, a lightweight community forum can help teams find expertise, but it rarely supports the deep back-and-forth needed to transfer tacit know-how. Evidence also suggests that effective transfer depends on both cohesion (trust and willingness to help) and range (access to non-redundant knowledge), which is exactly why “discovery” and “transfer” require different network designs (Reagans & McEvily, 2003).
3. Seed bridging intentionally
Once gaps and goals are clear, bridging should be designed, not left to chance. One of the most reliable ways to do this is to intentionally activate boundary spanning roles that sit between groups, because those positions are where novel information, translation, and coordination can happen (Burt, 1992; Cross et al., 2007).
A practical approach is to appoint a small number of boundary spanners per major gap, but with explicit time allocation and role clarity so bridging doesn’t become invisible labor. The underlying logic is that boundary spanning creates value by filling network gaps, but it also creates load; if the organization depends on one “hero connector,” it becomes fragile (Cross et al., 2007). Over time, the goal is not just to “have a broker,” but to increase the number of people who can bridge so coordination and translation become distributed rather than scarce (Burt, 1992).
4. Add scaffolding so bridges don’t collapse
Connections alone are fragile. Cross-silo collaboration holds when it’s supported by scaffolding: shared boundary objects (templates, roadmaps, metrics, prototypes) and explicit teaming norms that reduce friction. Research on knowledge boundaries shows that “boundary objects” are especially important because they help groups coordinate without requiring full agreement on language, assumptions, or priorities (Carlile, 2002).
This is also where psychological safety becomes operational rather than aspirational. Cross-boundary work involves risk: asking “basic” questions, exposing uncertainty, and negotiating conflicting constraints. Teams collaborate more effectively when members feel safe to speak up, learn, and experiment, conditions that have been empirically linked to learning behavior and performance in teams (Edmondson, 1999). In practice, scaffolding is what makes bridges repeatable rather than personality-dependent.
5. Measure and rebalance continuously
Effective silo reduction is observable in the network. Organizations should track increases in cross-silo ties, reductions in fragmentation (often reflected in lower modular separation between groups), and growth in cross-unit knowledge flows over time (Cross et al., 2007). If you have meeting metadata, this can be approximated by changes in cross-team meeting rate or the share of meetings with multi-team attendance, but the deeper signal is whether those cross-boundary ties persist and diversify rather than remaining one-off events.
Equally important: monitor broker load. When boundary spanners become bottlenecks, the solution isn’t to remove them, it’s to distribute bridging more evenly so the network becomes resilient rather than dependent. Practically, this looks like decreasing concentration of betweenness among a handful of people, improving the external–internal balance between groups, and reducing vulnerability to disruption if a single broker exits (Krackhardt & Stern, 1988; Cross et al., 2007).
Closing thought
Silos don’t disappear because leaders ask for collaboration, they dissolve when organizations intentionally reshape how connections form and are sustained. Network analytics turns silo reduction from a cultural aspiration into a measurable, operational capability.
References
Burt, R. S. (1992). Structural holes: The social structure of competition. Harvard University Press.
Carlile, P. R. (2002). A pragmatic view of knowledge and boundaries: Boundary objects in new product development. Organization Science, 13(4), 442–455. https://doi.org/10.1287/orsc.13.4.442.2953
Cross, R. L., Parise, S., & Weiss, L. M. (2007). The role of networks in organizational change. McKinsey Quarterly, (April web exclusive).
Edmondson, A. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350–383. https://doi.org/10.2307/2666999
Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 78(6), 1360–1380. https://doi.org/10.1086/225469
Hansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharing knowledge across organization subunits. Administrative Science Quarterly, 44(1), 82–111. https://doi.org/10.2307/2667032
Krackhardt, D., & Stern, R. N. (1988). Informal networks and organizational crises: An experimental simulation. Social Psychology Quarterly, 51(2), 123–140. https://doi.org/10.2307/2786835
Reagans, R., & McEvily, B. (2003). Network structure and knowledge transfer: The effects of cohesion and range. Administrative Science Quarterly, 48(2), 240–267. https://doi.org/10.2307/3556658
Interesting read!