---
title: "Chapter 16: The Birth and Death of an Outcome"
book: "Collapse or Collapse"
author: "Vinay Pasricha"
url: https://vinaypasricha.com/books/collapse-or-collapse/chapter-16-the-birth-and-death-of-an-outcome
status: "On Kindle: https://www.amazon.com/dp/B0HNL53J5R. The full text is shared on vinaypasricha.com with the author's permission."
---

# Chapter 16: The Birth and Death of an Outcome

## 1. One rule for both ends

The old company opened a division when a senior man wanted one and closed it when the board lost patience, and between those two events the division lived for years producing nothing anyone would have missed, because nobody had written down what it was for. The collapsed company has one rule for the birth and the death of an Outcome, and the rule is the test the book has used since chapter 5.

An Outcome exists exactly as long as a result has no other owner.

It is born the moment a result exists that nobody produces: a customer asks for something, or would pay for something, that no chain in the company is built to deliver, and no agent inside any vector can deliver as a step. It dies the moment that result has another owner: another Outcome’s vectors now produce it, as chapter 15 showed, or the customer who felt it is gone. Birth and death follow from the same question, which is the Keeper’s question in chapter 8: is this a result in its own right, and who produces it? When the answer is “a result, and nobody”, an Outcome is due. When the answer becomes “a result, and someone else”, the Outcome is done.

That is the whole law. Clayton Christensen spent a book showing that established firms fail at the new thing not because they cannot see it but because, inside the structure they have, it is nobody’s result;[^54] the rule above is the answer to him, applied every Monday.

The law has one exception, and it is a matter of capacity, not of result. When an Outcome’s floor grows past what one master can read, so that the reds reaching her week after week are more than her review can clear, the Keeper splits it in two, each whole, each with its own master and a copy of the playbook, as chapter 7 provides. Nothing new is produced and no result is unowned; the split exists so that no master is ever given a staff. The Keeper records a split as a new Outcome opened from the parent’s playbook, and the rest of this chapter applies to it. The rest of this chapter is procedure.

## 2. Who flags, who decides

Birth begins with a flag, and anyone with a name on a tile may raise it. The aligner who notices that a third of her enquiries ask for a service the company does not sell, the tuner whose agent maker keeps proposing a step the playbook has no place for, the master who sees the same unowned request on three vectors’ tiles: each may flag a result that nobody owns. The flag is a line in the record, not a presentation, and it says three things: what the result is, who would receive it, and why no existing chain can produce it as a step.

Only the Keeper decides a birth. That is not caution; it is the rule of chapter 8, that only he opens an Outcome, funds its first budget, names its first master and gives it its one result. A master cannot open an Outcome, because an Outcome is not inside hers. A pair cannot, because a pair holds a vector. The Keeper decides on the Monday test, in the record, within the week, and the answer is a birth or a reason.

Death needs no decision, because death is written on the wall. An Outcome whose result another Outcome now produces, or whose customer has gone, shows it on the Keeper’s table before anyone says so: its tile goes amber as its result thins, and the tile of the Outcome that has absorbed its work carries the result instead. The Keeper does not decide to close it. He records that it is closed, which is a different act, and section 7 shows it.

## 3. Born whole

An Outcome is born whole, or not at all.

This is the rule most likely to be broken by a well-meaning Keeper, so the chapter states it bluntly. A new Outcome has at least one master and at least one vector from its first day. It has a result of its own, a customer of its own, a playbook of its own and a Monday test of its own. It is not a vector that grew; it is not a trial that was promoted; it is not a project with a sponsor. A vector never becomes an Outcome, because a vector is one running copy of a discipline that already exists, and a new Outcome is a new discipline: a completely new kind of vector, built on a completely new thought about what the result is and how a chain produces it.

The distinction matters because the company already has a way of trying things, and it is not this. Chapter 8’s Keeper approved a trial vector for an overseas market, opened from the winning playbook, with a quarter to prove itself. That trial was a vector of the winning Outcome, aimed at a new market, and if it passed it became one more vector on that Outcome’s floor. Trials make vectors. They never make Outcomes, because a trial runs an existing playbook in a new place, and an Outcome needs a playbook that does not exist. When a flag arrives, the Keeper’s first question is therefore which of the two he is looking at: a known result in an unserved market, which is a trial, or an unowned result, which is a birth. The flag is often wrong about which it is, and the Keeper is the only person placed to tell.

## 4. Where the first master comes from

The first master of a new Outcome comes from inside the company, and the chapter rules out the alternative in one sentence: nobody outside the structure can hold a seat in it on the day it opens, because the seat is a craft learned on a floor, and an outsider has not stood on one.

So the first master is a backup master. Chapter 15 gave every master a backup who shares her control room and could hold it alone any morning; that is the person who has read a floor, settled an inspector’s appeal, kept a playbook and passed the four layers, and who is, by design, not indispensable where she sits. She takes the new Outcome. Her old seat takes a new backup, selected on intention, from the aligners, as chapter 14 describes.

The first vector’s pair are drawn the same way: an aligner or a tuner from another vector, who has held a chain and knows what a gate is, becomes the aligner of the new kind of vector, with a tuner selected to join her. The vector they leave takes a new pair through the four layers. Nothing in this is a promotion. A backup master who takes a birth is paid by the same formula on a result that begins at zero, and the doctrine expects her to earn less in the new Outcome’s first year than she did as a backup, which is one more reason the seat goes to the person who wants the craft.

## 5. What a death leaves

When an Outcome dies, two things happen to what it held, and the chapter will not soften either.

Its agents follow the work. The result the Outcome produced is now produced somewhere else, and the agents that produced it, the rubrics, the tools, the record of every decision they took, go into the vectors that now hold the result, where the agent maker fits them into a chain that already exists. Nothing an agent learned is lost, because an agent’s learning is in the playbook and the record, and both survive the Outcome.

Its people are offered any genuine vacancy they are suited to, and the rest are paid out. The pairs of a dead Outcome do not move to the absorbing Outcome, because the absorbing Outcome’s vectors already have their pairs and one human per seat is the doctrine. They do not move to the edge, because chapter 14’s rule about the edge is for pairs freed when vectors merge inside a living Outcome, where the company’s result is growing and new vectors are being opened; an Outcome’s death is the company’s result moving to fewer hands, not more. They are paid out as chapter 13 pays leavers, above what the law requires, on the day the record closes, and the record says why. The master of a dead Outcome becomes a backup if there is a master to back, as the master of Pravaha’s capital Outcome did in chapter 15, and is paid out if there is not.

A reader who finds this hard should notice that it is exactly what the old company did, after two years of pretending otherwise, to a division everyone knew was finished. The collapsed company does it on the day the wall says so, in writing, and pays better.

## 6. A birth and a death at Pravaha

The two happened in the same quarter, and the Keeper’s record shows them twelve lines apart.

**The birth.** The plant’s spare capacity had been a line in the fulfilment tiles for a year: machines cutting sixty-four per cent of the week, as chapter 13 left them, and the rest standing. The aligner of the OEM winning vector flagged it first, because three of her enquiries in a month had asked whether Pravaha would machine castings for other firms, and no chain in the company could say yes. Her flag said what the result was, machined components delivered to another manufacturer against its drawing; who received it, manufacturers who were not pump makers and would never be Pravaha’s customers for pumps; and why no existing chain could produce it as a step, because the fulfilment vectors’ agents were built to make Pravaha’s own units against Pravaha’s own order book and had no step for someone else’s drawing, someone else’s tolerance or someone else’s price.

The Keeper read it on the Monday. His first question was the one of section 3: known result, new market, or unowned result? A pump for a new customer is a trial. A machined part to another firm’s drawing is a different result, with a different customer who feels a different thing, and the chain that produces it would price, schedule and inspect against rules Pravaha did not have. A birth. He wrote the Outcome’s one result, spare-capacity-to-invoice, and its first budget, a quarter’s worth, in the record, and named its first master: the backup to the fulfilment master, who had shared the plant’s control room for eighteen months and had been the one, as it happened, who kept pointing at the idle thirty-six per cent. The OEM vector’s tuner became the new vector’s aligner, because she knew the castings and the machine shop; a tuner was selected through the four layers in the same week; the plant’s control room took a new backup from its aligners. The new Outcome had a master, a vector, a playbook written in the first fortnight from the fulfilment playbook with every rule about drawings, tolerances and pricing rewritten, and a Monday test of its own. It was born whole, and it invoiced its first casting in week seven.

**The death.** In the same quarter the design Outcome went. Design-to-first-unit had produced one result, the first unit of a new or changed design, for one customer, the plant. For two years the plant’s vectors had been drawing their own changes: their agent makers had absorbed the drawing and costing agents as a step inside the chain, because a change to a seal or a flange was a change to a unit the vector was already making, and the vector could draw it, cost it and cut it faster than it could hand the request across a boundary and wait. By the quarter in question, the design Outcome’s six vectors had produced one first unit in three months, and the plant’s vectors had produced eleven. The design tile had been amber for a year. The result had another owner. Nobody decided this; the wall said it, and the Keeper read the wall on a Monday and recorded it on the Tuesday.

The design Outcome’s agents went into the plant’s vectors, which already held most of them. Its master became the backup to the master of the new spare-capacity Outcome, which was a week old and needed one. Its twelve people, six pairs, were paid out on the Tuesday, each with a line in the record saying why, and two of them came through the four layers for the plant’s vectors inside the year, because selection for intention does not remember whose Outcome you were in.

## 7. The entry

The Keeper’s entry closing the design Outcome was one line, and it is reproduced here because the chapter’s argument is that it needs nothing more:

*Design-to-first-unit closed this day. Its result, a first unit of a new or changed design shipped against an order, is produced by the raw-material-to-finished-stock vectors, which have produced eleven to its one in the last quarter. Agents to those vectors. Master to backup, spare-capacity-to-invoice. Twelve seats paid out. No customer is affected.*

![Picture 14: Birth and death](https://vinaypasricha.com/assets/images/collapse-or-collapse/pictures/14-birth-and-death-1600.webp)

*Picture 14. Birth and death: the Keeper’s table with one tile going dark and one lighting, and the two entries twelve lines apart in the record.*

There was no announcement, because the people affected had read the same tile for a year and knew what amber meant. There was no farewell, because an Outcome is not a family; it is a chain built to produce one result, and the result was still being produced, better, by another chain. There was no inquiry into whether it should have been closed sooner, because the record shows the Monday on which the wall first said so, and the Keeper’s speed on reds is his own colour in the board’s view of him. The old company would have held a meeting. The collapsed company wrote a line.

## 8. Rules for the birth and death of an Outcome

1. An Outcome exists exactly as long as a result has no other owner.
2. Any human with a name on a tile may flag an unowned result. The flag states the result, the receiver, and why no existing chain can produce it as a step.
3. Only the Keeper decides a birth, on the Monday test, in the record, within the week.
4. Trials make vectors. Births make Outcomes. A known result in a new market is a trial; an unowned result is a birth. The Keeper tells them apart.
5. An Outcome is born whole: at least one master and one vector, a result, a customer, a playbook and a Monday test of its own, from day one.
6. A vector never becomes an Outcome. A new Outcome is a new kind of vector, built on a new thought.
7. The first master is a backup master from inside. Nobody outside the structure opens an Outcome.
8. The first pair are drawn from another vector; the seats they leave are refilled through the four layers.
9. A new master is paid by the same formula on a result that starts at zero.
10. Death is written on the wall: another Outcome produces the result, or the customer is gone. The Keeper records it; he does not decide it.
11. A dead Outcome’s agents follow the work into the vectors that now hold the result.
12. A dead Outcome’s people are laid off and paid out as chapter 13 pays leavers, on the day the record closes, with the reason written.
13. The master of a dead Outcome becomes a backup if there is a seat to back, and is paid out if there is not.
14. The closing is one line in the record. No meeting, no inquiry, no ceremony.
15. One birth is not caused by a new result: when an Outcome’s floor outgrows what one master can read, the Keeper splits it in two, each born whole, each with a master.

## 9. The birth worksheet

One page, filled by the person who flags and signed by the Keeper.

1. The result: one line, stated as what the receiver gets.
2. The receiver: who feels it, and that they are not already a customer of another Outcome for the same thing.
3. Why no chain produces it: the step that does not exist, in the nearest existing playbook.
4. Trial or birth: known result in a new market, or unowned result. The Keeper’s answer.
5. The first master: which backup, and who backs the seat she leaves.
6. The first pair: which aligner or tuner, from which vector, and who refills it.
7. The first budget: one quarter.
8. The playbook: which existing one it starts from, and the rules that must be rewritten before the first gate.
9. The Monday test: the one question the Outcome answers each week from week one.
10. The first invoice: the week by which the Outcome is expected to produce its result once.
11. The Keeper’s signature and the date.

[^54]: Clayton M. Christensen, *The Innovator’s Dilemma: When New Technologies Cause Great Firms to Fail* (Harvard Business School Press, 1997). Christensen’s finding was that established firms fail at the new thing not for want of seeing it but because it is nobody’s result inside the existing structure, and his remedy was a separate organisation with its own customers and its own measures. The collapsed company reaches the same remedy by rule rather than by exception: an unowned result is an Outcome, born whole, with its own master, customer and Monday test, on the day it is flagged.