DEEP READING · READ 935
Ameeba keiei · Kazuo Inamori · 2006 (with his 1998 book on management accounting, Jitsugaku)
Cut the company into several hundred units of five to fifty people; let each set its own monthly plan, buy and sell from the units next to it, and produce its own profit statement, so that everyone can see today whether their group made money or lost it — the point is not finer control, it is mass-producing a scarce species: turning one businessman into several hundred, until the shift supervisor on the line is doing the arithmetic too, and the arithmetic is his own.
Kazuo Inamori (1932–2022) founded Kyocera, later founded DDI (today's KDDI), and in 2010, aged seventy-eight and drawing no salary, took the chairmanship of Japan Airlines while it was in bankruptcy reorganisation. This book describes the divisional accounting system he worked out at Kyocera. It is the second half of the same man's project: his better-known A Compass to Fulfillment is the doctrine of character; this is the machinery that puts the doctrine on a report every single day. Management accounting scholars treat it as the canonical case of the micro-profit center — profit responsibility pushed down far below the level of a division.
An amoeba is a single-celled organism with no fixed shape: it deforms to fit its surroundings and, when it grows large enough, splits in two. Inamori chose the name for exactly those three properties — small, deformable, able to divide and recombine.
An amoeba is typically five to fifty people: a production line, a single process step, a sales team. It has its own leader, writes its own monthly plan, and keeps its own books. It is not an administrative department but a small business. When volume grows, one amoeba splits into two; when a line shrinks, two merge back into one. The org chart is not a poster revised once a year but a living body that divides and fuses as the market moves.
The origin is unglamorous: Kyocera grew faster than Inamori could personally supervise, and what he wanted was not more managers but more people who thought like an owner. The conventional answer is hiring and promotion — but the mindset of an owner is not taught, it is extracted by circumstance. Nobody genuinely cares about the electricity bill and the delivery date until they are responsible for a statement with their own name on it. So he did not set out to train businessmen; he set out to manufacture, in bulk, positions that require one, and then put ordinary employees into them.
The shift it forces: when you complain that your staff lack a sense of ownership, the truth is usually that your organisation contains no positions in which anyone could be an owner. Mindset is a product of position, not of training.
The heart of the system is a report called the hourly efficiency statement (jikan atari saisan). Its arithmetic is simple enough for a line supervisor who has never taken an accounting class:
| Line | What it contains |
|---|---|
| Total production | External sales plus sales made to other amoebas inside the company |
| Expenses | Materials, purchased parts, utilities, depreciation, floor space, purchases from other amoebas… but never the unit's own labour cost |
| Added value | Total production minus expenses |
| Hourly efficiency | Added value divided by the unit's total hours worked |
Two design decisions deserve a hard look.
First, wages are deliberately excluded from expenses. This is the line accountants most often attack, and the most carefully considered thing in the book. Inamori's reason is psychological: put payroll in a unit's cost line and every member, doing the daily numbers, is staring at his own salary as a deduction — being reminded each morning that he is a burden. Worse, amoebas would begin comparing whose people are cheaper, and start bidding labour costs down against each other. So he moved people out of the subtraction in the numerator and into the divisor: labour is no longer a cost but the time in which value is created. That single move changes the report's grammar from "how much did you spend" to "what is your hour worth."
Second, the divisor is hours, not headcount. Overtime therefore stops being free heroism: work two extra hours without a matching rise in added value and the number falls that same day. The metric closes off the easiest shortcut — throwing time at the problem — and pushes people toward yield, process and lead time instead.
The statement is updated daily, compared monthly against plan, and visible to everyone. Its real function is not appraisal but translation: it turns a vague collective goal into a question answerable before you go home — is our hour worth more today than it was yesterday? The shift it forces: most corporate metrics fail to move anyone because the feedback loop is too long and the attribution too muddy. Shortening the loop is itself an act of management.
The report alone is not enough. In a conventional manufacturer only sales has revenue; every upstream step is a cost centre — and the only contribution a cost centre can make is to spend less. It has defence available to it and no offence.
The decisive move is that amoebas buy and sell from one another. Forming sells semi-finished pieces to firing, firing sells to machining, machining sells to the sales amoeba, which sells to the customer and takes an agreed commission on the sale as its own revenue. Every internal handover is booked as an internal sale, so each process step suddenly has a top line, therefore its own profit and loss, therefore a business to answer for.
The power of this is that it reverses the direction of motivation. A cost-centre supervisor thinks "don't blow the budget." An amoeba leader with a top line starts thinking "can we lift yield and ship more," "can this batch go out two days early," "should we bring this step in-house." He has stopped being a man who saves money and started being a man who makes it.
The cost arrives in the same move: who sets the internal price? Set it high and the upstream unit is milking the downstream one; set it low and the upstream unit works for nothing. Inamori's answer is that internal prices must be derived backwards from the market — once the final selling price is fixed, it is apportioned back across the steps according to the value each adds, with disputes settled by the leader one level up. This is the most honest difficulty in the book: there is no objectively correct transfer price. It is permanently a negotiation requiring arbitration, and the arbiter's fairness is the load-bearing wall of the entire system.
Inamori repeats that the principle of business is almost banal: push sales as high as they will go, push costs as low as they will go, and what remains is profit. Manufacturers habitually assume that costs rise in proportion to volume; he refuses that. What he demands is revenue up with expenses flat — breaking the fixed linkage between the two. Amoeba management is the tool for breaking it, because the linkage only comes apart when every small unit is watching both numbers itself.
Its companion is the line he valued most: "pricing is management." A price is not a number quoted by the sales department; it is a decision the person running the business must make personally, because it simultaneously fixes three things — whether the customer buys, how much volume you get, and how much is left per unit. Price too high and the volume vanishes; price too low and a year's work yields nothing. His criterion is to find the single point at which the customer is glad to pay and your own profit is greatest — not the industry norm minus two points.
Why this matters outside manufacturing: it moves profit out of the finance department's year-end conclusion and into a decision made on the front line during every quotation. The shift it forces: most people assume profit is earned through work. Inamori points out that a very large share of it is settled in the thirty seconds when you name a price.
Inamori states three purposes, and the order matters:
The one most often left out of the copy is the second. Plenty of companies import only the first — push accounting downward, make appraisal more granular — and end up with an exquisitely detailed performance surveillance system instead of several hundred businesspeople. The distinguishing question is simple: is the statement a battle map the unit leader uses, or a report card his boss uses? The first makes people think. The second makes people manage the books.
His 1998 book on accounting is the other half of the machine, for the obvious reason that if the books are not trustworthy, several hundred unit statements are just several hundred elegant lies. He gives seven principles; four are worth memorising.
The remaining three are muscular management (no fat, no stockpiles, buy what you need when you need it), perfectionism (numbers may not be "roughly right"), and improving profitability (carrying the accounting consciousness into every transaction). The shift it forces: an incentive system's ceiling equals the credibility of the data under it. Where the books are soft, finer measurement is more dangerous, not less.
Divide a company into several hundred self-accounting units and parochialism follows almost by construction: upstream wants the internal price higher, downstream wants it lower, everyone wants to hand the unprofitable job to somebody else, and nobody wants to absorb a temporary reallocation for the good of the whole. Amoeba management is a very sharp knife: it amplifies the motive to answer for your own statement several hundredfold — and that motive is morally neutral.
Inamori's countermeasure is not another layer of rules but a shared criterion, drilled relentlessly: ask first what is right as a human being, rather than which choice flatters our unit's numbers. Kyocera wrote it down, recited it, and invoked it in meetings. He is explicit: amoeba management does not work without the philosophy.
This is also the part that transplants worst. A system can be copied; a criterion cannot. It has to be paid for daily in the conduct of the people at the top, and the moment senior management discounts it once in front of real money, the philosophy is demoted to a slogan in everyone's eyes — while the knife stays sharp. The shift it forces: any organisational design that pushes responsibility down and cuts units small must simultaneously answer who prevents local optimisation from eating the whole. The answer is either stronger central coordination or stronger shared values. It cannot be neither.
The book is one straight chain of reasoning: the principle of business is maximum revenue and minimum expense → but ten thousand people will not exert themselves for a ledger they cannot see → so cut the company into hundreds of units that keep their own books → use internal trade to give every process step a top line and make it a business → use one number, added value per hour, to make all units comparable and visible the same day → so that each unit leader is forced by circumstance into being an owner → and because that same knife amplifies parochialism, backstop it with a shared philosophy and books you can trust (one-to-one correspondence, the double check, glass walls, cash).
Its relation to A Compass to Fulfillment fits in a sentence: that book answers why you should raise your character; this one answers how, once raised, it shows up on a report every day.
1. The principle of business is one line — maximise revenue, minimise expenses. The hard part was never understanding it; it is getting ten thousand people to act on it at once.
2. So the company is cut into units of five to fifty people, each planning, accounting and reporting for itself: the org chart is not a poster but a living body that splits and merges with the market.
3. He did not want more managers, he wanted more people who think like owners — and that mindset is not taught, it is extracted by having to sign your own statement. Mindset is a product of position.
4. The one metric is added value per hour: (total production − expenses) ÷ hours worked — arithmetic a line supervisor can do in his head.
5. Wages are deliberately kept out of expenses and put into the divisor instead: people stop being a cost to be cut and become the time in which value is made; the report stops asking what you spent and starts asking what your hour is worth.
6. The divisor is hours, not heads, so overtime is no longer free heroism — put in two extra hours without more added value and the number drops the same day.
7. Let the units buy and sell from each other and every process step acquires a top line: a cost centre can only save money; a unit with revenue starts trying to make it. The price of that is a transfer price nobody can compute and someone must arbitrate.
8. "Pricing is management" — a price simultaneously decides whether they buy, how much you sell, and what is left per unit; a large share of your profit is settled in the thirty seconds when you name it.
9. Underneath sits the accounting: goods matched one-to-one with paper, every cash operation double-checked, numbers open to all, cash rather than accrual as the measure. The double check is not distrust; it is refusing to leave anyone alone with a temptation.
10. And the knife is neutral: amplifying "answer for your own statement" several hundredfold amplifies parochialism by the same factor. Hence Inamori's insistence that without a shared philosophy the system does not stand — every design that cuts responsibility small must say who stops local optimisation from eating the whole.