The people who make an organisation go are rarely the people the chart says are running it.
Why it matters when the plan changes
Most cross-boundary work in a large organisation runs on relationships rather than process, because the formal design rarely connects the places the work has to cross. Reading an organisation means looking past roles to real ownership and the dependencies a strategy relies on, and much of that ownership sits with people holding no formal mandate. That makes the informal network load-bearing and invisible. A restructure that removes a few of its nodes can disable coordination that no process document ever described.
The tension is that the network is both an asset and a risk. It is what makes the organisation work despite its structure, echoing what Rob Cross and Laurence Prusak called the people who make organisations go, and it concentrates dependency in people who were never given the role. This is consistent with findings that cross-unit commitments are the least reliable part of execution. Mapping the network makes the dependency visible and risks formalising something that works partly because it is informal.
In practice
A reorganisation removes two senior manager roles as duplicates. Both were the routes by which three functions actually coordinated. The structure after the change is cleaner and two dependencies that used to resolve in a phone call now take a fortnight, and no analysis of the change predicted it because the routes were never documented.
Evidence
Some individuals make organisations work through informal networks rather than formal authority.
Rob Cross and Laurence Prusak, The People Who Make Organizations Go or Stop, Harvard Business Review (2002)Cross-unit commitments are the least reliable part of execution, which is what the informal network compensates for.
Donald Sull, Rebecca Homkes and Charles Sull, Why Strategy Execution Unravels and What to Do About It, Harvard Business Review (2015)
What it cannot tell you
An informal network shows who people actually go to, not why the structure fails to connect them or what should replace those routes if the people involved leave. It cannot say whether the dependency is healthy or fragile, and mapping it captures a moment; the pattern can shift quickly once named or once a key node moves.
Questions
By asking who people actually go to for advice, decisions and information about what is happening, then noting who is named repeatedly across groups. This matters because research published in Harvard Business Review in 2015 found only 9% of managers could rely on colleagues in other functions and units all the time.
Because a restructure selects roles to remove on formal criteria and can remove the nodes that carry coordination. Rob Cross and Laurence Prusak described in Harvard Business Review in 2002 how such people make organisations go, and their loss shows up later as unexplained slowness between functions that no organisational chart predicted.
Sometimes, and carefully. Naming a connector into a role acknowledges what they do and can also remove the trust that made them effective, since people used them precisely because they were outside the hierarchy. Recognition and protection often work better than appointment.
It is a sign of adaptation. It means the organisation is compensating for structural gaps, which is admirable and also means the gaps exist. A network doing a great deal of work usually indicates a structure that does not connect the places the work needs to cross.
Concentration. Coordination that depends on a handful of relationships disappears when those people leave, change role or run out of capacity, and nothing formal replaces it because nothing formal was ever asked to. That is a key person risk that no role description shows.