The first org chart hangs in slide decks. Boxes, lines, titles. It gets updated after every reorganization, printed into onboarding documents, and consulted by almost nobody who actually needs something done.
The second org chart is invisible, and it runs the company. It's the map of who people actually call before committing to anything. Where proposals really get approved - usually before the meeting that officially approves them. Whose hesitation can stall an initiative without a single word of objection on record. Which one person's nod makes a "final" decision actually final.
Every company has both. The performance question is not whether the second chart exists, it always does - but how large the gap between the two has grown, and whether anyone has ever looked at it.
Why the second chart exists
It isn't pathology, and it isn't politics. It's physics.
Formal org charts encode reporting: who appraises whom, who approves vacation. They say almost nothing about deciding: who must be convinced, in what order, before anything real happens. And work, like water, routes around ambiguity. Where the formal structure leaves a question open - who decides pricing exceptions, who can commit engineering time, who settles a conflict between two departments - people find the actual answer through experience, and the actual answer becomes a path, and the path becomes the second chart.
It grows from legitimate material: competence ("she just knows this domain"), history ("he was here when we built it"), trust, and occasionally fear. Some of it is the organization at its best - expertise attracting influence regardless of title. Some of it is the organization at its most fragile - entire categories of decisions flowing through one unmapped person whose resignation would be an amnesia event.
What the gap costs
When the two charts diverge far enough, predictable things go wrong.
Initiatives stall in invisible approval loops. The project has formal sign-off, budget, a kickoff deck - and somehow doesn't move, because one informally load-bearing person hasn't been brought along, and everyone downstream senses it.
New leaders fail on schedule. They arrive, read the formal chart, and manage it - issuing decisions their title entitles them to make. The decisions don't stick.
Reorganizations redecorate. As I've argued elsewhere: moving boxes without moving decision rights changes nothing, because the boxes were never where the decisions lived.
And risk concentrates silently. The "bus factor" isn't only a software concept. Most companies have two or three people whose informal position is worth more than several formal departments - and no succession plan, because no plan can cover a role that officially doesn't exist.
Mapping it - without spying on anyone
The second chart sounds mystical. It's embarrassingly mappable, with methods a leadership team can run in two weeks.
- Trace five recent decisions end to end. Pick significant, recent, concrete ones - a pricing exception, a key hire, a tool purchase, a priority change. For each, reconstruct honestly: who raised it, who shaped it in private before any meeting, where it stalled and whose hesitation caused the stall, and who finally made it stick. Write the actual path next to the official process. The gap is now visible on paper.
- Ask the input question. Ask each leader, separately: "Whose input do you genuinely need before committing to something important - regardless of the org chart?" Draw the arrows. Hubs appear within an afternoon, and they're rarely where the formal chart predicts.
- Look at where work waits. Queues are confessions. Wherever decisions pile up waiting for one calendar, the second chart has a node the first chart doesn't show.
Notice what none of this requires: monitoring anyone, reading messages, or surveillance of any kind. You're mapping decision flows the way you'd map a supply chain - at the system level, from events everyone already knows happened.
In case you want a ready-to-go solution which will give you these answers and more - let me know! :)
What to do with the map
Three moves, in order of preference.
- Legitimize what works. Where the informal structure is effective - a person who has become the real decision-maker because they're genuinely the best one - give them the formal right. Aligning the charts beats fighting them, and it usually costs one announcement.
- Repair what hurts. Ghost vetoes - people who can stall decisions without accountability - get named into the formal process or removed from the path. Overloaded hubs get load-balanced by distributing the decision rights that secretly converged on them. Orphan zones - decisions belonging to nobody on either chart - get owners, in writing.
- Re-measure on a cadence. The second chart drifts. People join, leave, gain trust, lose it. A decision-flow review twice a year keeps the gap small; the work is only painful the first time.
The connecting thread
If you've read Decision Debt, you'll recognize where that debt physically accumulates: precisely in the gap between the two charts. Decisions that the formal structure assigns to nobody and the informal structure routes through somebody - unowned officially, bottlenecked actually - are the debt register and the shadow map of the same territory.
Which is the practical takeaway. You cannot manage the second org chart by pretending the first one is in charge. But you don't have to choose between them. Draw both, honestly. Then spend a quarter making them converge - and watch how many "people problems" turn out to have been geography all along.