The Silo Effect

Silos aren’t merely a communication failure. They emerge wherever organizations divide work into groups that develop their own expertise, relationships, and priorities, and the resulting boundaries are often protecting something real.

3 min read

What Is It?

“Silo mentality” has no single academic originator. It’s been part of management vocabulary since at least the 1980s, used to describe departments, teams, or systems that hold onto information and resources rather than sharing them across a boundary. Because the term has no one clean origin, it tends to get treated as a symptom of bad culture or bad incentives, something to be corrected with a mandate to “communicate more.”

Silos don’t have a single cause. They emerge when three ordinary organizational forces reinforce each other: scale divides relationships, specialization divides expertise, and incentives divide priorities.

Scale divides relationships first. Anthropologist Robin Dunbar’s research describes nested limits on stable social relationships, often summarized as circles of roughly 5, 15, 50, and 150, with the outer figure as the broadest commonly cited layer, not a demonstrated organizational threshold at which departments necessarily fracture into silos. As groups grow, personal familiarity can no longer carry coordination by itself, and organizations respond by dividing into smaller units where trust, language, and expectations stay manageable. Those units develop their own shorthand and their own sense of what good work looks like. The signal is audible before it’s measurable: listen for the moment a colleague becomes “they.” A twenty-person company rarely has one. A two-thousand-person one almost always does.

Specialization divides expertise next. Each cluster develops a distinct language and set of standards once it forms, the vocabulary and judgment that make it good at what it does.

Incentives divide priorities last, rewarding local performance more reliably than cross-boundary understanding and reinforcing the boundary scale first created. None of these mechanisms creates a silo alone, but together they make silos a predictable feature of organizational growth.

Gillian Tett’s The Silo Effect is the most substantial book-length treatment of the idea, tracing organizational silos to how humans classify and categorize information. Drawing on institutions caught in the 2008 crisis and several organizations outside finance, Tett showed how specialized groups can remain individually competent while the institution loses the ability to see across them. No single desk has to be incompetent for an organization to go blind. Someone just has to fail to be positioned to see across desks.

Why Does It Matter?

Most silo interventions start from the assumption that the boundary itself is the defect. But no one actually wants a legal team that reasons like a product team, or a finance team that thinks like marketing. Specialization is the entire point of drawing boundaries in the first place, and the same boundary that frustrates collaboration is usually the boundary that produced the expertise worth collaborating with.

That doesn’t mean every silo deserves protecting. A specialist team can be bounded without being closed, maintaining demanding internal standards while still publishing its decisions, rotating people through adjacent teams, and making its knowledge legible to outsiders. The same team turns dysfunctional when membership starts controlling access, information only moves inward, and decisions emerge from relationships no formal role can enter. One real version of the closed case: a leadership cluster that had been making the actual strategic decisions informally, outside any calendar, for nearly two decades. New executives, regardless of title or capability, couldn’t get in. Decisions arrived at the broader team already made, framed as recommendations but effectively final. The difference was never the existence of the boundary. It was whether anything could still cross it.

What Changes Once You See It?

You stop treating “the silos aren’t talking” as a discipline problem to correct with a mandate, and start asking whether this group has grown past the point where familiarity alone can coordinate it, and whether anything has actually been built to compensate.

You start listening for the pronoun. “They” about a group two floors away isn’t a personality quirk, it’s a tribal boundary becoming audible, and it tells you where to look before any engagement survey would.

You also start separating two different diagnoses that get treated as the same problem. A silo protecting real depth needs a bridge built around it, not dissolved. A silo that has turned into a gate, controlling what crosses and who benefits, needs its exclusivity made expensive, not more collaboration workshops aimed at people who were never actually the ones excluding anyone.

Common Misunderstandings

  • It is not purely a function of headcount. A ten-person team split across two shifts, two locations, or two competing priorities can silo just as thoroughly as a division of thousands. Scale makes a boundary more likely; it isn’t the only thing that draws one.
  • It is not an argument that all cross-team friction is healthy and should be left alone. Some boundaries really have turned extractive, and pretending every silo is secretly just “a professional community” is its own kind of blind spot.
  • It is not solved by increasing interaction alone. Asking people to collaborate more doesn’t automatically create shared understanding. Without changing how information, incentives, or relationships actually cross the boundary, more meetings often produce more coordination work rather than better coordination.
  • It is not the same phenomenon as bureaucracy or red tape. Silos are a social and cognitive limit, not a paperwork problem, and treating them with process fixes, more approvals, more forms, usually adds weight without addressing scale.
  • It does not mean an organization should try to stay small forever to avoid forming silos at all. Growth makes tribes inevitable past a certain point. The fix isn’t preventing scale, it’s building deliberate bridges once scale arrives.

Diagnostic Question

Where has “we” quietly become “they” in this organization, and has anything been deliberately built to bridge that boundary, or is it still running on trust that used to be automatic?

Explore Further

Field Notes

Related Field Guide

Origin

“Silo mentality” is a management term with no single academic originator, in wide use since at least the 1980s. Robin Dunbar’s research on the cognitive limits of stable relationships helps explain why organization-wide trust becomes harder to sustain as groups grow (see Dunbar’s Number). Gillian Tett’s The Silo Effect: The Peril of Expertise and the Promise of Breaking Down Barriers (2014 UK / September 2015 US) is the most significant book-length treatment of what happens next, tracing how specialization and classification then shape what organizations can and cannot see.

Know someone who’d enjoy this?