The bridge that keeps the banks apart
Some senior people describe themselves, with some pride, as the bridge between two parts of the company. Bridges, by definition, span a gap — and there is a well-documented incentive for the person standing in that gap to keep it open. Most of them don’t know they’re doing it.
There’s a particular kind of senior person who describes themselves, with some pride, as “the bridge.” The bridge between engineering and the founders. Between product and commercial. Between the people who build the thing and the people who decide what the thing should be. They carry messages back and forth, translate one side’s language into the other’s, smooth the friction. It sounds like exactly what a fracturing company needs.
It is often the thing keeping the fracture open.
The most quietly destructive role in a struggling organisation isn’t the obvious obstructionist. It’s the indispensable translator — the person whose value depends, structurally, on the two sides never quite learning to talk to each other directly. And the reason this is so hard to spot is that from the inside, and from above, it looks like help.
The mechanism is well understood
This isn’t a hunch. It rests on one of the more robust findings in organisational sociology: Ron Burt’s work on structural holes, developed over more than two decades.
A structural hole is a gap between two parts of a network that aren’t directly connected. Burt’s central finding, replicated across large studies of real organisations, is that the people who span those gaps — the brokers — capture disproportionate rewards. In his study of managers at a large electronics firm, compensation, positive evaluations, promotions, and credit for good ideas all clustered around the people whose networks bridged structural holes. Brokers were more likely to have their ideas heard and less likely to have them dismissed. Spanning a gap is genuinely valuable, both to the organisation and to the broker.
But look closely at why it’s valuable to the broker. Burt is explicit: the person standing in the gap gets disproportionate say over whose interests are served when the two sides are brought together. They control the flow. They frame each side to the other. They decide what gets relayed and what gets quietly dropped.
Now follow the incentive to its conclusion. If a broker’s value comes entirely from spanning a gap, then closing that gap destroys their value. The two banks, once joined, no longer need the bridge. Burt has a paper titled, plainly, “Reinforced Structural Holes” — the broker has a structural incentive to keep the parties apart, and the structure tends to reproduce itself. This is, in the language of organisational economics, a principal–agent problem: the agent’s private interest (remain essential) diverges from the principal’s interest (an integrated, well-coordinated organisation). The silo isn’t an accident of communication. It’s the equilibrium the broker’s incentives push towards.
It’s usually not malicious
This is the part that matters most, because the obvious reading — “the bridge is a manipulator deliberately keeping people apart to protect their job” — is usually wrong, and believing it will make you misdiagnose the situation.
Most bridges are not cynical. They very often believe, sincerely, that they are the integrating force in the company. “I’ll go and talk to them and come back to you” feels like help. Running a meeting with one side, then carrying its conclusions to the other, feels like diplomacy. The role feels integrative from the inside at precisely the moment it is preventing integration. That’s what makes it a principal–agent problem rather than simple bad faith: the agent doesn’t know they’re misaligned. The incentive operates underneath their intentions, not through them.
I’ve watched this happen. An engineering organisation and a founder, with one senior person shuttling between them — meetings with one side, then the other, never the two in the same room. Meanwhile a commercial leader was coming to the technical team in genuine distress, asking how the company could be brought together to solve problems everyone could see. The bridge talked constantly about bridging. The integration never came. Not because anyone was a villain, but because the only person positioned to convene both sides was the one whose role quietly depended on them staying separate.
How to see it
The pattern is invisible until you know what you’re looking for. Then it’s everywhere. The clearest signals:
Your senior team doesn’t actually meet as a team. If the SLT holds its conversations in separate rooms — one person sitting in one meeting and then the other, acting as a relay — you are looking directly at the bridge, and you can usually see both banks on either side of it. A leadership team that genuinely functions as a team comes together, regularly and directly. Persistent separation is the tell.
Information between two groups flows through one person rather than between the groups. Ask yourself: if engineering needs something from the founders, does engineering talk to the founders, or does it talk to the bridge?
One side’s position is always framed by the broker, never voiced by the side itself. You hear “what the business really wants is…” or “engineering’s concern is…” far more than you hear those groups say it themselves.
Attempts to put both sides in one room don’t happen, or meet quiet resistance. This is the strongest signal. If convening the two groups directly is somehow always difficult, ask who finds it difficult and why.
One caution, because the diagnosis is easy to over-apply: not all separation is a problem. Temporary working groups are healthy. A small team that breaks off to go deep on something and then rejoins the wider org — an iceberg that calves and drifts back to the mainland — is doing exactly what it should. The pathology isn’t periodic, purposeful separation. It’s permanent, total separation, where two parts of one company exist in steady-state isolation with a single human relay between them. If the islands never reconnect, that’s the thing to get to the bottom of.
The test that tells good bridges from bad
Here’s the cleanest diagnostic, and it works even though the surface behaviour of a healthy broker and a structural-hole broker can look identical.
A healthy bridge is working to make itself unnecessary. It actively connects the two sides to each other — introductions, shared forums, direct lines — so that over time they need the go-between less. A structural-hole broker, consciously or not, works to remain essential: the connections always route back through them.
So the question isn’t “is this person carrying messages between two groups?” Plenty of legitimate roles do. The question is directional: is this person’s activity increasing or decreasing the amount of direct contact between the two sides over time? A good bridge’s success metric is its own obsolescence. A bad bridge’s instinct is to preserve the gap that gives it standing.
If you suspect you might be the bridge, the uncomfortable version of the same test: would you be comfortable making yourself redundant by connecting these two groups directly? If the honest answer has a flinch in it, that’s information.
What to do about it
If you’re a leader and you can see this in your org, the fix is structural, not interpersonal — you don’t need to remove the bridge, you need to remove the gap.
Put the two sides in the same room, regularly, with a shared artefact between them: one backlog, one set of objectives, one board everyone reads from. Shared objects force direct contact in a way that no amount of goodwill does. Create lateral channels that don’t route through a single person. And specifically engineer the bridge out of the information path — not as punishment, but because as long as the path runs through one person, the incentive to keep the banks apart will keep reasserting itself, regardless of anyone’s intentions.
The deeper point is that an organisation’s structure has its own incentives, and those incentives will quietly defeat people’s good intentions if you let them. The bridge who keeps the banks apart isn’t a bad person. They’re a predictable output of a structure that rewards the gap. Fix the structure and you don’t need the bridge — which is, after all, what a bridge is supposed to want.
This is the second in a short series on the structural patterns that quietly break technical organisations. The first looked at why companies keep hiring people who can’t save them. Next: what founders actually mean when they ask for a “product team” — and why it’s rarely what they get.