Org chart

The design of organisations and products The Jolly Contrarian holds forth™ →Go uptown intro sa 💥🆕💥 Audio version of this article 🎧 Listen to this podcast — on Spotify — on Substack — JC Life on Apple Index: Click ᐅ to expand: Audio Essays Resources and Navigation Index: Click ᐅ to expand: Design Index: Click ᐅ to expand: Essays

What you see is all there is.

— Daniel Kahneman

Der Teufel mag im Detail stecken, aber Gott steckt in den Lücken.
“The Devil may be in the detail, but God is in the gaps.”

—Büchstein, Die Schweizer Heulsuse

Org chart
/ɔːg ʧɑːt/ (n.)

A formal portrait. A still life. A glib schematic that tells you everything you don’t need to know about an organisation, but which it treats as its most utmost secret.

The org chart purports to order the organisation, placing everyone in a fixed, hierarchical relation to everyone else and joining them with reporting lines that radiate out and down from the the splayed fingers of the chief executive. Therefore, a centrally-sanctioned, aspirational, blueprint: to the executive suite, what the “built environment” is to the town planner: a superficially plausible account of how the organisation is meant to work, that bears no relation to how it does work.

The plan you have before you get punched in the mouth.

Form, not substance

Organisations are organisms. They have a way of frustrating their executives’ best-laid plans, just as cities — also organic, self-organising systems — delight in upsetting politicians and urban planners’ platonic aspirations.

This is no accident, but a necessary consequence of forward motion into an uncertain future. An org chart is static; an organisation must move. A firm that devoutly obeys its present operating manual is, to all intents and purposes, on strike — that is the literal definition of a “work-to-rule”.

For an organisation is what it does, not what it is. Org charts say as much about what an organisation does as an ordinance survey map does about the weather in a given location, or how people will behave if it rains. Being static, they speak to what is meant to happen in an expected future that behaves according to the historical model. They do not accommodate contingencies, opportunities, and unexposed risks. They contain only the vertical communication channels that personnel are meant to use, to respect the firm’s governance structure, not the lateral ones they must use to move the organisation forward, much less the informal ones they do use, because they want to, and because — to hell with the rules — these desire lines have proven the best way to get anything done.

Desire lines
/dɪˈzaɪə/ /laɪnz/ (n.)
The informal pathways we make when our “built environment” lets us down. They cross open ground, emerging over time as individuals make their own judgments as to the best and most convenient routes, often ignoring central planning and its designed pathways.

Desire lines, yesterday. © 2014 Steve Bates

We should not underestimate the importance of the “want” in that calculus, by the way: there is always a choice whom to call to progress a given task. All other things being equal, we choose those who we have found to be helpful, co-operative and imaginative over those who tend to be defensive, hostile, boring or stupid. It is reflexive: “no good deed goes unpunished”: over time, popular staff field more calls, get more experience, build better networks and give better outcomes: “want” converges with “need”. Lesson: if you want to get ahead, don’t be a dork.

In any case, these vital informal communication channels rarely run along the formal lines of the org chart. Why would they?

What management sees is all there is

Management is obliged to focus on this formal, static structure, made flesh in reporting lines, because that is all it sees. Consider an imaginary employee: Dan Grade.

Dan is an ED in the risk team. He is an agreeable chap: he handles anything from dumb questions to tricky escalations with equanimity. He’s runs a tight ship during his twenty-five years in organisation building a trove of institutional knowledge and personal capital which he applies deftly to manage risk, tamp down flare-ups, calm intemperate traders and head off incipient snafus, patiently educating generations of grads, juniors and his own line management in the mystical ways of sound of risk management, while maintaining a heroic sense of composure, a superhuman tolerance for tedium, time wasting and the petty initiatives of middle management.

Dan does all this so well that he rarely comes to the attention of anyone important. No news is good news. From the Eagle’s Nest, he’s just tiny, fungible node in the thickets of branches fanning out across the valleys and plains of the org chart chart below. Only his reporting line and salary is visible from the executive suite. They can count and optimise the spans and layers of which he is part, and attribute to them the profits and losses of the organisation even if, in practice, they don’t map awfully well, but they have no clue as to what Dan does to move the organisation on.

Closer at hand, there are hundreds people — all of them indistinguishable nodes on the org chart, of course — who know the place would fall apart without people like Dan.

Management can’t see how many people rely Dan in practice: only those who do in theory. Management only sees the most sclerotic, rusty and resented communication channels in the organisation: the organisation’s reporting lines: these are the “keep off the grass” signs; vain attempts to coerce inferior modes of communication over better ones for, if reporting lines really were the best lines of communication, you would not need to formalise them: they would just happen, the same way lateral communications just happen to flow into Dan.

One-to-ones and the formal business of reporting

Since they don’t, management exhorts line managers to meet weekly with their directs, populating standing agendas to furnish management information and statistics fit for injection into opco decks and RAG dashboards of handsome looking but, given the circumstances of their generation, basically useless data.

Now this is not to suggest that there are no meaningful communications between line manager and direct report: that would be absurd. If they are both in the office, they will be constantly be in contact — one more reason working from home is not the paradigm shift some would like to believe — relaying important information to each other in real time. But this, too, will be an informal dialogue: unminuted, off-the-record, oral, instantly evaporating, plausibly deniable and, until formalised, entirely beyond management’s understanding.

To formalise that live dialogue into “management information” it must first be filtered, to take out the usual interpersonal pleasantries, low-level exchanges about technical details and updates about the BAU: the “weeds”. What should remain are revenue, cost and risk items, the last of which being the hitches, snafus, market dislocations and brewing ructions with clients and counterparties that are the basic raison d’être of a risk department.

Here, a key objective is to get this information to the boss P.D.Q., so she isn’t blindsided by someone further up the chain. If someone outside your chain of command knows about an unfolding problem, you best make damn sure your manager knows it, so she can tell her manager about it, and so on, up the chain. But all this, to reiterate, is still informal, off-grid communication.

By the time an unfolding “situation” makes it into the formal written record, it has been euphemised, contextualised, narratised and put into an absolving passive that will give little hint of the potential enormity of the problem, whilst still allowing scope to claim “I did tell you” should the problem reach its potential.

Thus, formal communications along the official chain of command are as likely to mislead management into a state of complacency as they are to alert it to a real problem. This was more or less the tale of the Credit Suisse losses on Archegos. Management never knew what hit it.

To be sure, the information revolution has made the business of frankness in internal communications ever more fraught. Anything put in the electronic record is discoverable, both in principle and practice. That regulation and technology should discourage open and early communication of risks as a matter for some regret.

When the firm is in motion

So reporting lines are a static map of the firm, configured in the abstract, when it is at rest. That is, before it does anything. This is how the machine works when it is idling. The firm’s business only gets done when its gears are engaged. And that happens when its personnel communicate with those who are outside their immediate reporting line.

This is the firm’s real arterial network: the lateral interactions that must cross whatever boundaries management can dream up, or that leave the firm altogether: these are the communications that employees must make: between internal specialists in different departments; with the firm’s clients and external suppliers — they make commerce happen and move the organisation along.

The org chart doesn’t say what should happen if Dan from risk needs to speak quickly to Janice in legal. The organogram says Dan must escalate up three layers to the Chief Risk Officer who will then speak to the General Counsel, who will “cascade” his thoughts down to Janice. Of course, that is not how it works, ever: Dan just picks up the phone to Janice, or vice versa. This is a desire line, worn by decades of communal habit.

These desire lines are pervasive across the organisation. It is in these interactions that things happen: it is here that tensions manifest themselves, problems emerge and opportunities arise, and here that these things are resolved.

The Dan Grades and Janice Hendersons are the ad hoc mechanics who keep the eighteen-wheeler on the road. They are the “super-nodes”: they know histories, have networks and understand intuitively how the organisation works, what you must do and to whom you must speak to get things done. They may not be especially “senior” — they are too busy to be senior — and they don’t derive their significance from their formal status, but from their informal function.

A bottom-up map of functional interactions would disregard the artificial cascade of formal authority in favour of informal credibility. It would reveal the organisation as a point-to-point multi-nodal network, far richer than the flimsy frame indicated by the org chart. With modern data analytics, it would not even be hard to do: Log the firm’s communication records for data to see where those communications go: who chats with whom? who calls whom? Who emails whom? What is the informal structure of the firm? Who are the major nodes?

Modernism vs. agilism

Management sees the firm as a centrally-managed, unitary machine that must be controlled from the top: the more structure the better. This is a modernist view. The alternative view is to see it as a self-organising ecosystem that does what it does in spite of management. That view advocates removing layers, disassembling silos and decluttering the structure.

If risks and opportunities arise unexpectedly, in times and at places you can’t anticipate, the optimal organising principle therefore is: give talented people flexibility and discretion to react as they see fit. Have the best people, with the best equipment, in the best place to react skilfully. Those people aren’t middle managers, the optimal equipment isn’t the org chart, and that place is not the board room, steering committee or the operating committee.

It is out there in the jungle. A confident management should trust its highly-paid specialists and seek to minimise the formal impediments to the creative behaviour of those people. It should put in place systems which attract, recognise, incentivise and develop the value of those loyal specialists.

So to understand a business one needs not understand its formal structure, but its informal structure: not the roles but the people who fill them: who are the key people whom others go to to help get things done; to break through logjams, to ensure the management is on side?

These lines will not show up in any org chart.

See also

• Line management
• Reduction in force
• Dotted line
• Spans and layers