The informal org chart
A Socratic walk-through of the informal org chart — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why is the person everyone asks for help rarely near the top of the chart?
Every workplace has a chart on the wall and a different chart in everyone's head. The one on the wall says who reports to whom. The one in people's heads says who you actually go to when the machine stops at half past four and the job has to ship in the morning.
The second chart has a striking feature: the person at its centre is very often a chargehand, a long-serving technician, a scheduler — somebody several boxes down. Nobody designed that. Nobody voted. So where does a stable, widely agreed structure come from when no one made it?
Reasoning it through
REASONING #Start with the smallest unit: one person, stuck, deciding whom to ask. What are they weighing?
Three things, roughly. Whether the person can actually answer. How long it will take to get the answer. And what asking costs them — in time, in interruption, and in what it reveals about their own competence.
That third term is heavier than it looks. Asking is an admission, and who you admit it to matters. Asking your manager puts your gap in the mind of the person who writes your appraisal. Asking the old hand two benches over does not. So even when the manager knows the answer perfectly well, the cheaper route is sideways or downward.
Now add the second term, speed. A senior person is, almost by definition, in meetings. Availability falls as you go up. And the knowledge that resolves a stuck machine at half past four is usually specific and local — this machine, this customer, this workaround — which is exactly the knowledge that accumulates through years of doing the work and tends to dissipate as people move up and away from it.
So for each individual asker, the local calculation points away from the top. Fine — but that only explains one person's choice. The question was about structure. Take the next step and ask what happens when the same calculation runs a few hundred times.
Suppose the answers are good. The asker gets unstuck, and now knows something new: that this person is worth asking. They ask again. They also mention it to a colleague — "go and see Denise." So each successful question makes the next question to the same person more likely, both from that asker and from others. That is a positive feedback: being asked produces being asked.
And it compounds twice over, because answering is itself a way of learning. Someone who fields thirty problems a month sees thirty situations; the more they are asked, the more they know, which makes them a better answer than they were. Their advantage is partly manufactured by the asking.
Watch what that does to the distribution. If everyone chose whom to ask at random, requests would spread out roughly evenly. With feedback of this kind, small early differences — a bit more experience, a manner that does not make you feel stupid, a desk on the way to the kettle — get amplified into large final ones. You do not end up with a flat spread; you end up with a few heavy concentrations and a long tail of people hardly asked at all. Nobody chose that shape. It falls out of many individually sensible choices, each made with no view of the whole. That is what emergence means here: a structure at the level of the organisation that exists in no one's intention and in no one's head, produced entirely by local rules.
I should be careful about one thing. The general finding that help-seeking networks are lopsided and only loosely aligned to the formal chart is well established in organisational network research. The precise shape of that lopsidedness — how heavy the tail is, whether it follows any particular law — is contested and varies by workplace, so I will not claim more than "a few hubs, a long tail."
The analogy
THE ANALOGY #Think of a path worn across a lawn. The council laid paved routes at right angles; the grass shows a diagonal. No one convened to route it — each person simply took the shortest line to the gate, and every crossing killed a little more grass, which made the next person's route slightly easier, which killed a little more. The path is a real, stable, load-bearing structure that emerged from footsteps and encodes information the designers did not have: where people were actually going.
the desire path costs its walkers nothing and asks nothing of the grass, whereas the hub of a help network is a person absorbing a real, uncompensated workload — so the informal structure can wear its own path out in a way a lawn cannot.
Clarifying the model
THE MODEL #Three refinements.
First, the informal chart is not a rival to the formal one; it is doing a different job. The formal chart allocates authority and accountability — who may commit money, who answers for a failure. The informal one routes knowledge. They are only in conflict when someone mistakes one for the other, which is what has happened when a manager treats being bypassed as a challenge to their standing rather than as a map of where the expertise sits.
Second, the hub is unpaid and invisible. The cost of being everyone's first call lands as interruption, and it lands on someone whose measured output is their own work, not the twenty problems they unblocked. This is why the hub is often the person who looks stretched and slightly behind, and why the harm of losing them is discovered only after they have gone — the graph reroutes badly, and nothing in the accounts predicted it.
Third, and least comfortable: because the structure emerged rather than being chosen, it optimises for what the local rule actually rewarded, which was cheap and quick, not correct. Emergent does not mean wise. A hub can be central because they are approachable and confident while being wrong more often than the specialist nobody wants to bother.
What would refute the account? If help requests tracked formal rank in workplaces where asking carried no reputational cost — a genuinely blame-free shop — then status cost, not availability or local knowledge, is doing most of the work.
A picture of it
THE PICTURE #How to readThe left-hand nodes are where questions come from, the right-hand nodes where they go, and each ribbon's width is how many travel that way. The point is the imbalance on the right: the line manager, highest on the wall chart, catches one of the thinnest bands, while a person with nobody reporting to them absorbs the widest. Note too that the flows cross departments freely — these routes were chosen one question at a time, not laid down by anyone.
What became clearer
WHAT CLEARED #The informal chart is not a rebellion, a rumour mill, or a failure of the real one. It is what you get when hundreds of people separately pick the cheapest, fastest, least embarrassing route to an answer, and each successful route makes itself more likely next time. That feedback concentrates the traffic on a handful of people whose position on the wall chart is irrelevant to why they are there — which is why the structure is stable, why nobody designed it, and why the organisation's real dependencies are invisible in its own accounts.
Where to go next
ONWARD #- How to find the hubs deliberately, and why simply asking "who do you go to?" changes the answer once people know it will be read.
- Whether formalising a hub — promoting them, making the role official — preserves the network or destroys the very approachability that created it.
Key terms
TERMS #| Term | What it means |
|---|---|
| Emergence | structure at the level of a whole system that arises from local rules and is present in no participant's intention or view. |
| Hub | a node receiving a disproportionate share of a network's traffic; here, the person a large fraction of help requests converge on. |
Every term the collection defines is gathered in the glossary.