---
title: "Chapter 10: The Twelve Tenets"
book: "Collapse or Collapse"
author: "Vinay Pasricha"
url: https://vinaypasricha.com/books/collapse-or-collapse/chapter-10-the-twelve-tenets
status: "On Kindle: https://www.amazon.com/dp/B0HNL53J5R. The full text is shared on vinaypasricha.com with the author's permission."
---

# Chapter 10: The Twelve Tenets

A doctrine is what is left when the stories have been told and the reader must decide what to do on Monday.

The nine chapters before this one were stories, some of them mine and some of them the century’s: the reorder book, the department as a workaround, the man who switched off, the motor in the basement, the cockpit at nine in the morning, the floor of tiles, the Keeper’s quiet room, the vector drawn where the customer feels it. Stories persuade. They do not instruct. A reader can agree with every one of them and go back to a company that is organised exactly as it was, because agreement costs nothing and a rule costs something. This chapter is the rules.

There are twelve. I did not set out to have twelve; I set out to write down every principle the collapsed company cannot do without, and to throw away every one it could. Twelve is what survived. Each is stated in one line, which is the line to remember. Each then has three short parts: why it holds, which is the constraint it answers; what it replaces, which is the practice of the old company it ends; and a test, one question an owner can ask of his own company on Monday morning and answer honestly by lunch. If you can pass the twelve tests, your company has collapsed, whatever it calls itself. If you cannot, it has not, whatever it calls itself.

One distinction first, because the earlier chapters stated some things as flatly as these twelve, and they are not of the same kind. A tenet is a principle: it holds in every company that collapses, and a company that breaks it has not collapsed. A setting is a number the doctrine gives as a starting value and the playbook changes on the evidence: how long a reversible gate waits before it opens, when a vector may hold more than its pair, how many reds a master can hold before her Outcome splits, the minimum standard a candidate must meet on each value, the threshold of judgement at which a new aligner signs alone. Where this book has given a setting a single number, it is the number I would start with, not a law. Change a setting when the record says so, and write down why. Change a tenet and you have built a department again.

## I. The department was a workaround.

**Why it holds.** Every function in every company exists because no single person could hold every skill an outcome needs. The forecaster could not also negotiate; the negotiator could not also draft the contract; so the outcome was cut into skills, and each skill was given a desk, and the desks were gathered into departments, and a manager was put above each department to carry the purpose the desks could not see. That was an honest answer to a real constraint, and it lasted two hundred years. The constraint has gone. A chain of agents holds every skill the outcome needs, in one system, without a handoff. When the constraint goes, the workaround has no reason left to exist.

**What it replaces.** The organisation chart drawn in boxes by skill: sales, operations, finance, HR, legal, technology. The belief that these boxes are the company, rather than scaffolding put up around it.

**The test.** Pick any department in your company and ask what result it produces that a customer would notice if it stopped. If the honest answer is “it helps the others produce theirs”, it is a workaround, and its work belongs inside the chains of the results it helps.

## II. Organise around outcomes, not activities. Work that produces a result is an Outcome; work that serves a result is an agent.

**Why it holds.** An outcome is a change the receiver feels: a mandate signed, a candidate who joins, a patient discharged, an invoice paid. An activity is a step towards one. The old company organised around activities because activities were what its people were good at; the collapsed company organises around outcomes because outcomes are what its customers pay for. The second sentence of the tenet is the one that keeps the first one true. Whenever a piece of work is found that does not obviously belong anywhere, ask whether it produces a result in its own right, one that someone receives and would miss, or whether it is a check or a skill used in producing someone else’s. Raising capital from the market is a result; it is an Outcome, with its own specialised vectors and a master who signed up to deliver it. Compliance is a skill. So is the finance check on a price, the legal review of a contract, the quality inspection of a batch, the screening of a hire. Each is an agent, standing inside the chain of every vector that needs it, at the step where it is needed. The same word, finance, can name an agent in one company and an Outcome in another. The test is never the name of the work; it is the outcome wanted.

**What it replaces.** The function. And, more dangerously, the function at the top: the legal head, the finance head, the head of people, sitting beside the chief executive with a hold over every result and responsibility for none.

**The test.** List every person who reports to the top of your company. For each, ask: what Outcome does she master? If the answer is a function rather than a result, the department has come back in through the front door.

## III. Every outcome runs on a vector.

**Why it holds.** A result is produced by something, and in the collapsed company that something is a vector: one result, aimed at one market, carried by one chain of agents, and held by one pair of humans by default. The aligner runs it and answers for what it delivers. The tuner builds and upgrades it and can run it alone. The vector is the smallest unit that produces anything a customer would notice, and nothing larger is needed to produce it. An Outcome is many vectors running one playbook; a company is many Outcomes.

**What it replaces.** The team, which was a group of humans dividing a result between them, each dependent on work he could not see. And the process, which was the department seen from the side: a sequence of steps across desks.

**The test.** For each result your company sells, can you point to the one chain that produces it end to end, and name the two people who hold it? If the result passes through three departments on its way to the customer, it has no vector yet.

## IV. Machines hold the baseline; humans supply the peaks.

**Why it holds.** Agents carry continuous work to a measured standard. They do not tire at the end of a shift, lose interest on a Friday, or switch off in the month before they resign. Humans do all of these, and also do what no agent can yet do: apply intention, urgency and judgement at the moments where the result turns. A company that asks humans to carry the baseline wastes them on the part of the work they do worst, and starves the peaks of the attention they need. The collapsed company gives the baseline to the chain and spends its humans only where a human changes the result: at the gates, at the signature, in the daily review.

**What it replaces.** The working day as a measure of contribution. The belief that more hours at a desk produce more result.

**The test.** For one vector, or one department if you have no vectors yet, count the hours its humans spent last week on work an agent could have carried to standard. Then count the hours they spent at the moments where only a human could have changed the result. The first number should be falling every quarter. If it is not, the machines are not holding the baseline; your people are.

## V. Pay for intention, not skill.

**Why it holds.** Skill is now cheap. The chain carries it, and the chain’s cost is a fraction of one salary. What is rare, and has always been rare, is intention that does not waver: the commitment to the result that holds on a bad week, in the month before the bonus, after the third rejection. One aligner with that intention, and a chain beneath her, produces what a department produced. She is worth to the company something like what the department was worth. So the doctrine pays her very highly, far above the market rate for her role. The level is deliberately high, for two reasons. At that price the company can choose from the very few whose intention holds. And at that output the price is still cheap, because the company needs a tenth of the people it needed before.

**What it replaces.** The salary band set by skill and seniority, and the old company’s habit of calling its people its greatest asset and paying them like a cost.

**The test.** Ask what you would pay the one person whose intention you would trust with a whole result, and whether you pay it. Then be honest about why not. GoodSpace does not yet pay at that level, and I will say why: its vectors do not yet produce enough to carry it. The design of chapter 9 is twenty-five mandates closed for every matched pair every month, and GoodSpace’s delivery vectors close eight to twelve. The pay follows the output; it does not lead it. But the principle is the right one, and a company that does not set a very high level of pay as its target will never find out what intention costs, or what it is worth.

## VI. No silent stalls.

**Why it holds.** Ford’s moving line had one property that the office never copied: when a station stopped, the line stopped, and everyone could see it. The office was built of silent stalls, files sitting on desks for a week while every report sent upward remained true. A vector cannot work that way. Every agent reports its own flow, all the time, and if any agent produces nothing for longer than the playbook allows, the quartermaster raises the alarm and the tile turns amber where the master can see it. A human gate that waits too long is the same. A vector that can stall silently is not a vector. It is a department with software.

**What it replaces.** The status meeting, the weekly report, the escalation, and the customer’s phone call asking where her order is. Each was a way of discovering a stall after it had cost something.

**The test.** Find the oldest piece of unfinished work in your company today. How long has it been waiting, and who knew before you asked? If nobody knew, you have silent stalls, and every report you received this week was true and useless.

## VII. Every vector has a standby.

**Why it holds.** People are ill, take leave, have children, and leave. The old company treated absence as an interruption and handled it by panic: the work waited, or a colleague who did not know it picked it up badly. The collapsed company designs the standby in. The tuner who builds the chain knows it end to end, and takes over when the aligner is away; the aligner takes over when the tuner is. Another master stands by for a master. A board member stands by for the Keeper. Gates that are reversible default open after a period the playbook sets, usually a day, so a missed day costs the vector one day of inspection, not one day of flow.

**What it replaces.** The handover note, the acting manager, the hiring in a crisis, and the single point of failure that every company has and few will name.

**The test.** Name the person in your company whose two-week absence would stop a result. Then name the person who would run it in her place, and ask whether that person could start tomorrow without a handover. A vector held by one person has no standby, and will discover it on the day she is ill.

## VIII. Count vectors, not employees.

**Why it holds.** The old company measured itself in people because people were its unit of production; it did more by employing more. The collapsed company’s unit of production is the vector, and so its size is the number of results it carries at once and the human intention behind each. The question at the top changes. The Keeper does not ask how many people he needs. He asks how many vectors the market will support, and how many each master can hold. This is span of outcomes, not span of control, and the design runs into the hundreds; when a floor holds more than its master can read, the Outcome splits.

**What it replaces.** Headcount as a measure of growth, of status and of a manager’s importance. The annual plan written as a hiring plan.

**The test.** Describe your company in one sentence without mentioning how many people work in it. If you cannot, you are still counting the wrong thing.

## IX. The aligner’s clock is her own.

**Why it holds.** The agents work round the clock, in the customer’s time zone. The aligner works in hers: where her attention is highest, when it is highest, for a number of hours she fixes in advance and then keeps. She chooses the hours; she does not choose whether to keep them, because the vector and the master need to know when its human will be there. For the first time since the factory bell, the worker’s clock is decoupled from the customer’s. The customer is served at three in the morning by the chain; the aligner reads what the chain did at nine, at the hour she chose, with her full attention.

**What it replaces.** Office hours, the shift, the attendance register as a proxy for work, and the commute as the price of being seen.

**The test.** Ask your best people when in the day their attention is highest, and then check whether that is when they do the work that needs it. If their best hours are spent in meetings and their judgement is applied at six in the evening, you are paying for intention and spending it on the clock.

## X. Hire for intention; measure attention.

**Why it holds.** Intention is the commitment to the result. It is selected for once, carefully, at a high price, because it is very hard to create and very easy to lose. Attention is that commitment showing up day by day, and it varies; everyone has a few bad days, and the ones with intention bounce back. When intention goes, the signs are written on the wall long before the resignation: the numbers dip, the leave increases, the person becomes unresponsive. The old company saw this late or never, because it watched people rather than results. The vector measures attention continuously, through the flow of the result: the reviews held, the gates answered, the batches sampled. The engine sees it first.

**What it replaces.** The annual appraisal, the attendance register, the manager who watches, and the exit interview that discovers what everyone already knew.

**The test.** Think of the last person who left your company of her own accord. How many months before her resignation could the data have told you she had switched off? If the answer is “none, we were surprised”, you are not measuring attention. You are measuring presence.

## XI. Measure, don’t price.

**Why it holds.** Outcomes serve each other without paying each other. The winning vector hands its mandate to the delivery vector, and neither sends the other an invoice. There is no transfer price, no internal profit and loss per Outcome, no internal market. Each Outcome has a real budget, held by its master, and real payments to the people and agencies who run its vectors; but between Outcomes, value is measured and not priced. The command centre sees what each Outcome delivers and what it costs, and the Keeper moves money between them on that evidence. A firm that prices every handoff becomes mercenary at every handoff, and spends its intention negotiating with itself.

**What it replaces.** The cost centre, the internal recharge, the transfer price and the profit centre, and the year-end argument over who earned what.

**The test.** Count the hours your managers spent last quarter negotiating budgets, recharges or credit with other parts of your own company. Every one of those hours was spent on a customer who does not exist.

## XII. A vector is a vector, whoever pays it.

**Why it holds.** A vector may be run by an employee, a contractor, a freelancer, a part-timer, a fractional pair or an outside agency, and several of these may run vectors inside one Outcome at the same time. A master may be fractional; in time the Keeper may be an agency. The form of payment makes nobody internal or external. The cockpit, the control room and the command centre do. Anyone who works from a cockpit, is read by the control room and is seen by the command centre is inside the company, and anyone who is not, is not. Membership is the screen, not the payroll.

**What it replaces.** The boundary of the firm drawn by the employment contract, and the distinction between staff and vendors that made half of every company’s work invisible to its own management.

**The test.** List every result your company depends on that is produced, wholly or partly, by someone you do not employ. Can you see each of them on the same screen as your own people, with the same alarms? If not, part of your company is outside the company, and you will learn about its stalls from your customers.

## Where GoodSpace holds them

I have used GoodSpace throughout this book as proof, and I have tried to be honest about where the proof is thin. Here I will name only the three tenets it holds most fully, because a doctrine is better served by three tenets held well than by twelve held in part. The tenets it holds partly or not yet are named, each where it belongs, in the chapters that discuss them.

| Tenet | Where GoodSpace holds it |
|---|---|
| **II. Organise around outcomes, not activities** | Held in the operating company: winning mandates is one Outcome, and delivering a mandate from job description to invoice is another, mastered by Mudassir. Neither is a department. Broken at the top, as the next section admits. |
| **IV. Machines hold the baseline; humans supply the peaks** | Agents search, assess and interview candidates around the clock; the humans’ half day is spent at the three gates, on the findability report, the client questionnaire and the personal sign-off on every candidate sent. |
| **VIII. Count vectors, not employees** | The company describes itself as twenty vectors across four Outcomes, each held by one or two people, and plans its growth as vectors opened, not people hired. |

Three of twelve, two of them held completely and one held below and broken above. It is less than I would like and more than any company I knew ten years ago could have claimed. The other nine are the work of the next few years, and the gaps are already on the tiles.

## The tenet that is broken first

One tenet is broken more often than any other, and it is broken quietly, by people with good intentions, usually within a year of the collapse. It is the second.

It does not happen through the chains. A vector that has been drawn around a result tends to stay drawn; its alarms, its tiles and its customers keep it honest. It happens at the top. A company collapses its operations into vectors and Outcomes, and then, because some kinds of work feel too important or too risky to leave inside a chain, it keeps one or two functions beside the Keeper. Legal stays there, because the contracts matter. Finance stays there, because the money matters. People stays there, because hiring matters. Each one is reasonable. Each is staffed by someone senior and trusted. And each, quietly, acquires what every function at the top has always acquired: a hold over every result, responsibility for none, and a reason to grow.

Within two years the hold becomes a queue, the queue becomes a team, the team becomes a department, and the department has a seat at the table where the Outcomes are decided. The old company has walked back in, through the front door, wearing the doctrine’s own clothes.

I know this because I have done it. As I described in chapter 8, I hold GoodSpace’s legal and finance myself, from the board, with our outside accountant. It was reasonable. It is also a function at the top, and it breaks the tenet I have asked you to hold above the others. The fix is not to find a better person to hold it. The fix is to ask the question of Tenet II and act on the answer. The legal review of a client’s contract is a check, so it belongs as an agent inside the winning chain, at the gate before the signature. The finance check on a price is a skill, so it belongs as an agent at the step where the price is set. Raising capital, if we do it, is a result, so it will be an Outcome, with its own vectors and a master who has signed up to deliver it. None of them is a seat beside the Keeper.

So watch the top. Once a quarter, list everyone who answers to the Keeper, and for each ask the question this chapter began with: what Outcome does she master? Where the answer is a function, the department has come back, and the collapse has begun to uncollapse. Put the work where it belongs, as an agent in a chain or as an Outcome with its own vectors, and empty the seat.

The other eleven tenets can be held badly for a while and recovered. The second, once it is lost at the top, takes the rest with it.