The Org Chart Is a Lie: A Field Guide to Reading Actual Power

The org chart was on the third slide, and it was beautiful.

Clean boxes. Solid lines descending in orderly cascades, dotted lines swooping in to indicate "collaboration," the whole thing rendered in the company's two brand colors like a diagram of a nervous system that had never once experienced pain. Forty-one boxes. A tidy hierarchy of accountability, printed on recycled stock, handed to me in a folder at 9:04 a.m. on my first day along with a laptop and a water bottle nobody asked for.

By 2 p.m. on day three I understood that the chart was, in the strictest sense, fiction. Not a lie exactly — every box was real, every name spelled right. It was fiction the way a subway map is fiction: technically accurate about connections, wildly wrong about distance, and completely silent on the question of which trains actually run.

Here's what I mean. There was a woman in that org chart — call her the Director of Platform Operations — sitting four levels down a branch nobody in leadership seemed to think about much. Small team. Unglamorous mandate. And in eleven weeks of watching, I never once saw a significant decision get made without her having already blessed it, days earlier, in a conversation that appeared on no calendar and produced no artifact.

She was not in the meeting. She had simply made the meeting's outcome inevitable before it started.

That's the second org chart. Every company has one. It is never printed, never discussed at onboarding, and it is the only one that decides what happens to you.

Why the gap exists (and why it isn't anyone's fault)¶

The tempting read here is cynical: the chart is propaganda, real power is hidden, everyone is scheming. That's wrong, and worse, it's lazy — it'll make you paranoid instead of effective, which is a bad trade at any price.

The truth is more structural and more interesting. A formal hierarchy is a compression algorithm. It takes a chaotic, high-dimensional reality — hundreds of people, thousands of dependencies, a tangle of who-knows-what — and squeezes it into a tree, because trees are the only structure a human executive can hold in their head and a payroll system can process. Compression is lossy. That's not a bug in the concept of an org chart; that's what compression is.

What gets lost in the squeeze is exactly the stuff that determines outcomes: who trusts whom, who owes whom, who has been here long enough to know why the fence is there, who can get a decision unstuck with one message on a Thursday. None of that fits in a box on a slide. So it goes somewhere else. It goes into the informal structure, which forms automatically, without anyone deciding to form it, because the work has to get done and the official diagram doesn't describe how.

Jo Freeman nailed this in 1972, writing about ostensibly leaderless activist groups, and the argument has aged into something close to a law of nature:

“

There is no such thing as a structureless group. Any group of people of whatever nature coming together for any length of time, for any purpose, will inevitably structure itself in some fashion.

”

Freeman's real point is the sharp one, and I'd tattoo it on the forehead of every founder who says the word "flat" in a job posting: a structure you refuse to name doesn't disappear — it just becomes unaccountable. "Flat" is doing exactly what the rest of corporate speak does — describing a thing by naming its opposite and daring you to argue. When there's no official hierarchy, there's still a hierarchy; it's just invisible, unelected, and impossible to appeal to. The flattest company I ever worked inside had the most rigid power structure I have ever encountered, and its rigidity was load-bearing on the fiction that it didn't exist. You cannot file a complaint about a hierarchy that leadership has declared imaginary.

Conway's Law: the product is a confession¶

In April 1968, a computer scientist named Melvin Conway published a paper in Datamation after the ACM rejected it, which is its own small comedy. It contained this:

“

Organizations which design systems are constrained to produce designs which are copies of the communication structures of these organizations.

”

This is one of those observations that is either trivial or profound depending on how long you sit with it. Sit with it.

It means the thing a company builds is physical evidence of who was talking to whom. Not who was supposed to be talking. Who actually was. Every seam in a product is a place where two groups of humans had to coordinate and coordination is expensive. Every baffling handoff, every feature that works beautifully right up until it touches another feature, every "you'll need to contact a different department for that" — those aren't engineering failures. They're org chart fossils. You're looking at the shape of a conversation that didn't happen, preserved in the product like a footprint in ash.

This is the single most useful diagnostic tool in this entire article, and you can run it on day one, before you've met anybody:

I have never once found this to be wrong. Two teams that hate each other ship two products with a bad seam between them. Two teams that trust each other ship one product. The customer experiences the emotional state of an organization they will never meet, and calls it "the app being annoying."

It also runs upstream of the org itself. How a company was funded shapes who it hires, which shapes who talks to whom, which shapes what it ships — which is why bootstrapped and venture-backed companies produce visibly different products long before anyone writes a strategy deck about it.

The three roles that actually hold leverage¶

After enough years of watching this, the informal chart resolves into a small number of recurring positions. None of them correspond reliably to title. All of them are occupied by someone right now at your company.

The Gate¶

The Gate is whoever can say no and make it stick without escalation. Not the person with formal veto power — the person whose objection is treated as terminal because overriding them is more expensive than the thing you wanted.

The Gate is very often mid-level. Sometimes it's a staff engineer with twelve years of context. Sometimes it's the head of a compliance function whose "I have concerns" ends discussions that a VP's enthusiasm couldn't restart. Find the Gate by watching what dies. Track five initiatives that had executive support and quietly evaporated anyway. Ask who raised the first serious objection to each. The name that recurs is the Gate, and you should have coffee with them before you need anything.

The Translator¶

The Translator sits between two groups that cannot understand each other and converts meaning across the boundary. Sales and engineering. Clinical and product. Legal and everybody.

Translators accumulate power for a reason so simple it's almost funny: they are the only path between two necessary things. Every conversation that has to cross that boundary crosses through them, which means they see more of the organization than anyone above them and get to frame how each side understands the other. They are usually underpaid, frequently exhausted, and their departure causes damage that takes leadership two quarters to attribute correctly.

The Institutional Memory¶

The one who knows why. Not what the policy is — why the policy is. Why the third-party integration everyone hates cannot simply be ripped out. Which client relationship the Tuesday ritual is protecting. Why the last three attempts at the obvious fix ended badly.

This is Chesterton's fence as a full-time job. G.K. Chesterton's rule was that you don't remove a fence until you know why it was put there; the Institutional Memory is the only person who knows, and they are typically not senior, and they are typically ignored right up until the fence comes down and everyone discovers what it was fencing.

What this does to your career, concretely¶

Here's the part that pays for the reading time.

Almost every "my great work went nowhere" story I have heard — and in the trenches you hear a lot of them, usually around the fourth drink, usually with more hurt in it than the teller intends to show — is a story about a person who executed flawlessly against the formal chart. They did the work. They did it well. They routed it to their manager, who routed it up, and it entered a system that had no idea it was coming, was not prepared to want it, and had three people in it who would have needed to be consulted first and weren't.

The work was fine. The work was never the variable. They optimized for the diagram in the folder.

And on the other side: the person who seems to move through the company like it's frictionless, whose projects mysteriously get resourced, who is somehow never blindsided. It is very tempting to call that person political and go back to being right and stuck. Sometimes that's accurate! Sometimes they're a courtier and the company is worse for them.

But usually — genuinely, in my experience, usually — they've just read the second chart. They talked to the Gate in week one. They know which of their VP's yeses are real. They pre-socialized the idea with the Translator so that when it hit the meeting it arrived already understood. That's not manipulation. That's the professional equivalent of reading the room, and refusing to do it isn't integrity, it's just an expensive way to feel righteous.

The Peter Principle, and why seniority is a bad proxy¶

One more thing, because it explains a phenomenon that otherwise makes new people feel insane.

In 1969, Laurence J. Peter and Raymond Hull published The Peter Principle, whose central claim — half joke, half devastating — is that in a hierarchy, employees tend to rise to their level of incompetence. You get promoted for being good at your job, repeatedly, until you land in a job you're not good at. And there you stay, because there's no further promotion and nobody demotes.

Whether it's rigorously true is arguable and researchers have argued it. What's definitely true is the corollary, which is the useful part: position in the formal hierarchy is a record of past performance in a different job. It tells you what someone was good at before. It does not tell you what they're good at now, and it certainly doesn't tell you whether they can help you.

So stop using title as a proxy for capability, or for power. The senior title tells you about compensation and blame-routing. It does not tell you who can get your thing unstuck on a Thursday afternoon. Frequently that person is four levels down a branch nobody thinks about, with a small team and an unglamorous mandate, and she has already decided what's going to happen in the meeting you haven't been invited to.

Frequently asked questions¶

The folder¶

I kept the folder from that first day. It's in a box somewhere — laptop returned, water bottle lost, org chart still crisp on its recycled stock, forty-one boxes in two brand colors.

Every name on it was correct. Every reporting line was accurate. It described, faithfully and in good faith, a company that did not exist — a company where decisions flowed down solid lines and information flowed up them, where the boxes at the top knew things and the boxes at the bottom did things, and where a Director of Platform Operations four levels down a quiet branch would have been exactly as important as her box suggested.

The real company was running on trust, debt, memory, and a few people who could say no. It always is. Every one of them is.

Learn the second chart. Nobody is going to hand it to you in a folder. It's just business — and business is people, arranged in a shape that nobody drew.

Sources

1. Melvin E. Conway, 'How Do Committees Invent?' — Datamation, April 1968
2. Jo Freeman, 'The Tyranny of Structurelessness' (1972)
3. Laurence J. Peter & Raymond Hull, The Peter Principle (1969) — Encyclopaedia Britannica
4. G.K. Chesterton, The Thing (1929) — origin of 'Chesterton's Fence'