Chapter 099　What Should New Entrants and Incumbents Redefine?

Prescription for Incumbents The question is one. What should be redefined? But it has two addressees. The enterprise that holds nothing yet, and the enterprise that already holds a great deal. The first has nothing accumulating. The second cannot move what has accumulated. Against a single question, the two answers are not symmetrical. This chapter treats that asymmetry as its structure. As the working conclusion of Vols. IX and X, we narrow to that one point.

1 Why this question arises now

The eighteen chapters from 081 to 098 asked what was redefined — the work of reading something already done across five dimensions. Chapters 099 and 100 ask what should be redefined: the work of giving order to something not yet done. The grounds for a prescription come out of those eighteen. What has to be extracted is not scale. It is the structural lesson. And a prescription does not work without an addressee. By new entrant we mean an enterprise born in the Age of AI, holding neither an existing business to defend nor organizational inertia. There is a good model, a working product, and a first customer. All three can be in place and the business still fail to survive. By incumbent we mean an enterprise that has lasted a long time, operates at scale, is often listed, and carries structures built for a different environment. Two answers have circulated about this form for years. It can no longer move, says the first. It has strengths no one else has, says the second. Neither connects to tomorrow’s decision. The difference between the two is not a difference of capability. The problem sits on opposite sides. For the enterprise that holds nothing, the problem is that speed produces no accumulation. For the enterprise that holds much, the problem is that what has accumulated does not move. What the first needs is a design for accumulation. What the second needs is a measurement of its range of motion. AI is pressing both at once. Foundation models can be connected and rented, and computing capacity can be bought by the unit consumed (→ Vol. VIII, Ch. 072). But the same conditions are handed to founders everywhere at the same time. That entry has become easy means that entry keeps happening. On the incumbent’s side, analytical capacity that once belonged to large enterprises alone is now available at any scale. The point of difference moves from holding the capability to deciding what it is for. The change of premise is common to both. What incumbents have been good at is precision after the objective has been handed down. Quality, unit cost, delivery date. This is also the territory AI enters fastest. When competition on precision levels out, what remains is the setting of the objective. So this question arises now not as a phase of the business cycle, but as a phase in which the definition of the enterprise is reopened.

2 Conventional answers and their limits

Each of the two addressees has its standard arguments. All are partly right, and all stop short of action. The conventional answers on the side that holds nothing “Build it and ship it. Speed is everything.” But speed carries no direction, and it is no longer a source of differentiation either. What made development fast was the evolution of the tools, and competitors use the same tools. Speed that accumulates nothing is only the speed at which fuel is spent. “Put a thin layer for a specific workflow on top of a foundation model.” This is the most dangerous conventional answer in circulation. Development is light and early revenue appears. The problem is whose decision makes that layer disappear. Providers of foundation models absorb, as standard features, the functions that many users ask for in common. The moment a function is absorbed, the value of that layer approaches zero. What decides is not your company but the other party’s development plan. A state in which your survival depends on one line of another company’s roadmap. That is the precise meaning of riding thin. Thinness is not a shortage of features. It is that nothing accumulates even as the product is used. The description of the product reduces to a clever prompt. Cancel the contract and the customer is not inconvenienced. After a year, nothing remains inside that a competitor could not reproduce. Where those three hold, the business is thin. The conventional answers on the side that holds much “Enterprises that have lasted a long time cannot change.” The evidence offered is slow decision-making, deference to precedent, and inward-facing personnel systems. These are observed, but the explanation is weak. First, under the same institutions and the same markets, some enterprises have rewritten the definition of their business. If institutions were the cause, every company in that market would land in the same place. Second, the view treats culture as a cause. But culture is usually the result of institutions and incentives. People choose precedent in large numbers because a mechanism rewards it. Mechanisms can be changed. “Enterprises that have lasted a long time have technology and shop-floor capability.” That these capabilities sit at a high level is confirmable across many industries. But a strength has conditions of use. Shop-floor capability is the ability to realize a given objective with high precision. In the language of the Enterprise Redefinition Maturity Model (ERMM), it is the capability that perfects an Improvement Enterprise. The canon describes that level this way: organizations at this level become increasingly efficient while remaining fundamentally unchanged. And AI is coming to carry part of the capability to improve a process. The enterprises that differentiated on the speed of improvement are the ones whose source of difference thins out. What all four answers miss All four stay inside an argument about what to build and what you hold. What does the enterprise exist for? How much of the value it creates stays with it? Over what time does it accumulate? All three sit outside the field of view. And each pictures change as pressure applied from outside. But the Continuous Redefinition Enterprise in the ERMM redesigns itself before external disruption requires it. The question is not whether you change. It is when you decide to, on your own.

3 How the answer splits across the five dimensions

We set the two answers side by side along the five dimensions of Enterprise Redefinition. The order follows the canon. What follows describes general tendencies and does not hold for every company. Purpose — a blank, and a poster Most new entrants did not begin from Purpose. They began from what this technology can do. That is not a bad starting point, but it gives the business no outline. Purpose → Learning → Redefinition → Creation → Enterprise Value Enterprise Value comes last. A company that began from technology is running the later terms with the first one blank. The blank is examined when a strong person is hired, when capital comes in, and at the first large failure (→ Vol. VIII, Ch. 072). Without a criterion for what to keep and what to discard, a change of direction cannot be distinguished from drift. For the incumbent it is the reverse. The creed exists, and the older the company, the more carefully it has been handed down. The problem is connection. Most such statements are declarations about being a good company, expressions of an attitude such as integrity or contribution. They do not state which societal challenge the enterprise takes on. Purpose works as a criterion only when some proposal is rejected because of it. So the work runs in opposite directions. The side that holds nothing fills a blank. The side that holds much connects the creed to the budget. Neither has rewriting as its object. The canon holds that Core Purpose can remain stable while its expression and its means of realization evolve. The incumbent needs the same core, restated in the language of a present societal challenge. The new entrant needs an answer to one question: what change would we still want to leave behind if our technology became obsolete? The core must not contain the name of a technology. Purpose Precedes Profit. Business — an outline too rigid, or no outline at all The canon does not ask what a firm sells. It asks what value the firm delivers. Change the unit and the competitors change, the required capabilities change, and the destination of capital changes. The incumbent has too much outline. Most describe themselves by product name, by industry name, or by the classification their trade association uses. Trade channels and long-standing exclusive networks of counterparties have held that self-definition in place. Japan is where the structure is most sharply visible, and there it has a name: keiretsu, the durable groupings of suppliers, customers, and affiliated firms, often reinforced by cross-shareholdings. Once the route is taken as given, the delivered value never has to be put into words. Value not put into words cannot be priced. Terms of trade get set by the counterparty’s logic because the seller holds no unit of value. The new entrant has too little outline. Value creation and value capture are independent variables (→ Vol. VII, Ch. 069). Creation can be large while capture is close to zero. The model and the computing capacity are both borrowed, and a substantial share of the value added flows upstream. Ease of adoption is, unchanged, ease of cancellation. Capture is therefore not an outcome but an object of design. What accumulates with use and cannot be reproduced quickly by others? At which layer of the value chain is the unit of billing placed? Put it at the lowest layer and creation and capture diverge sharply (theory → Vol. VII, Ch. 069; a worked case → Vol. X, Ch. 094). What is opened, and what is closed? Will the firm own the errors that come with AI output? A business that owns them cannot be a thin layer, and the record of having owned them is the asset hardest to imitate. Organization — relationships that are old, or relationships that do not exist In the canon, an organization is not a collection of people. It is a value-creation system made of people, AI, partners, universities, and customers. In many incumbents the org chart is a map of where people sit. AI does not appear on it. Neither do suppliers, universities, or customers. What is not on the chart does not reach the agenda. There is, however, an overlooked asset here. A network of long-term trading relationships — keiretsu in the Japanese case — is accumulated relationship across the boundary of the firm, and the raw material of an ecosystem already exists. What is missing is not the relationships. It is a design that treats them as a value-creation system. Long-term trading is assembled to stabilize supply. An ecosystem is assembled to create new value. The new entrant is the reverse. The purpose is new; the relationships do not exist. Customers, holders of data, the people who actually do the work, the parties to regulation. A design for interaction with these does not appear in the business plan. Capital — stagnation, and a maturity date In the canon, capital is not only financial capital. It includes knowledge, brand, trust, AI, networks, and data. In the incumbent, the capital allocation meeting has two items: plant and acquisitions. Spending on people, knowledge, data, and trust is treated as expense rather than investment. Expense is something to cut, not something to allocate. The difference in treatment is often explained as a matter of accounting standards, but the management accounting frame is something an enterprise designs for itself. In the new entrant, the problem is not quantity but character. Investment has a maturity, and the period to realization is set by the structure of the capital rather than by the needs of the business (→ Vol. VIII, Ch. 072). Generations of foundation models turn over on a shorter cycle than that. If the horizon for return runs longer than the life of the technology, the premises change before what should accumulate has accumulated. If it runs shorter, the cycle on which results are demanded falls below the cycle of learning: Learning drops first, and then Redefinition stops. Capital is chosen by its maturity, not by its amount. First Principle 3 states, Capital Exists to Create Possibility. On one side, capital that creates no possibility sits still. Sitting still is not safety; it is a choice to lose value slowly. On the other side, only the deadline has been brought in. The symptoms are opposite. The principle is one. Leadership — approval, and self-recognition In many incumbents the executive’s job is approval. The canon defines management differently: a move from individual judgments to the design of the future. Approval answers a question someone else has framed. Management frames the question. The gap is not one of ability but of the allocation of time. And AI is rapidly taking over the work of preparing the material for approval. What the freed time is spent on determines whether the leadership dimension is redefined. Human-on-the-Loop Management does not ask for supervision of outputs. It asks for design of the whole system. In the new entrant, the issue is not time but self-recognition. Founders often still think of themselves as engineers. From the moment capital is accepted, that person stands in the position of designing a future. And the machinery that protects a purpose can only be designed before capital comes in (→ Vol. IX, Ch. 085; Vol. X, Ch. 098). Articles of incorporation, voting structures, and restrictions on the use of funds cannot be moved afterward.

4 Structure — which term is closest to zero

We apply the equations and look at the skeleton. The point is that they are products. If any single term is zero, the whole product is zero. No term compensates for another.

4.1 The place that goes to zero is reversed

Value = Purpose × Trust × Capability × Time In the new entrant, Capability is high. The technology is real and the product works. But Purpose is close to blank, Trust has accumulated nothing, and Time runs on someone else’s clock. If three terms are near zero, raising Capability does not enlarge the product. The technology is not what is missing. What is missing is something to multiply it by. The incumbent is the reverse. Future Capital = Financial × Human × Learning × Trust × AI × Knowledge × Ecosystem × Purpose Long-term trading relationships, quality guarantees, and delivery discipline. That this accumulation has thickened Trust is observable across many industries. Trust Compounds Faster Than Capital. But most of that trust has accumulated with existing counterparties and has not necessarily been transferred to new partners or new domains. What is thick is the balance, not the flow. Knowledge that has not been made explicit cannot be loaded into AI either. What drives the term is not total volume but circulability. Purpose is not zero; what is low is its effective value. And maturity is not an index of scale. Small companies can be Reactive Enterprises too.

4.2 What they share is a thin Capital Reallocation Capability

Enterprise Redefinition Capability comprises six capabilities: Strategic Intelligence, Learning Capability, Design Capability, Capital Reallocation Capability, Leadership Capability, and AI Collaboration Capability. We hold that the fourth is thin in both forms alike. In the incumbent, the evidence is in how budgets are set. Where the method is to argue increments against last year’s actuals, reallocation structurally cannot occur. This is not an absence of capability. It is a consequence of procedure. The Capital dimension of the ERMM asks: “Are resources allocated toward Future Value rather than historical success?” A budget built from last year’s numbers is, by definition, allocated toward historical success. In the new entrant, the evidence is in what the budget contains. FVCC = Purpose × Learning × Redefinition × AI Integration × Ecosystem × Capital Allocation × Trust AI Integration is high from the start. But the allocation debate leans toward headcount and advertising, and no allocation runs toward assets that accumulate. That Ecosystem is low we saw in the previous section. With two terms near zero, the product is small — weakness in any single capability weakens the whole.

4.3 The Leadership Formula and the term of time

Leadership = Purpose × Question Design × Capital Allocation × System Architecture × Trust In the incumbent, the term most likely to approach zero is Question Design. The machinery of circulated approval is a finely built apparatus for examining answers, but it has no step for framing a question. In Japan the practice carries a precise name: ringi, the procedure by which a written proposal circulates upward, collecting the seals of each responsible party before it takes effect. The underlying mechanism of accumulating approvals from below is common to large enterprises everywhere. In the new entrant this term is high instead, because founding is itself a question. Future Value = Future Time × Future Capability Enterprises that have lasted a long time are often called longterm oriented. But long-term orientation and Future Horizon are not the same thing. An orientation toward continuing the existing business extends the horizon backward, into the past. Horizons are also easily trimmed to the length of an executive’s tenure. The new entrant’s strength lies, properly, in the length of Future Time. But that time is shortened by expectations borrowed in advance. If the inflection point of the technology coincides with the point at which funding runs out, the most important judgment gets made at the moment of least bargaining power. The same term moves toward zero for opposite reasons.

4.4 What is evaluated is coherence

The ERMM evaluates organizational coherence rather than isolated excellence. Progression across levels is not linear: an organization can hold Level 4 AI capability while remaining Level 2 in leadership, and among new entrants that shape is the most common of all. Maturity is also assessed across all five dimensions in balance; exceptional technological capability with weak leadership redesign does not produce higher maturity, and strong purpose without adaptive organizational systems remains insufficient. And Level 5 is not a target to be reached as fast as possible, because different industries require different levels of adaptability.

5 What it looks like in practice — constraint or excuse

We break the claim that an enterprise cannot change into five structures common to companies that have lasted a long time. For each we separate the part that institutions genuinely fix from the part that sits inside management’s discretion. This is the core of the chapter.

5.1 Employment practice

The object is the personnel system built on long-tenure employment and seniority-weighted pay. It survives most sharply in Japan, but long employment guarantees and treatment linked to length of service exist, to some degree, in large companies in every market. What cannot be moved. The law and case law governing dismissal are not something a company changes on its own. A business shift premised on headcount reduction cannot be executed at the speed available in other markets. What can be moved. The same legal framework grants wide discretion over reassignment. You cannot cut people, but you can move them. Moving talent between businesses is, legally, comparatively easy. That it rarely happens is because business units treat people as owned resources. That is an internal practice, and moving where personnel authority sits changes it tomorrow.

5.2 Shareholder composition

The object is a register of shareholders fixed by the history of relationships, of which cross-shareholding is the emblem. Cross-shareholding appears most sharply in Japan, but a structure in which founding families and lending banks hold shares for long periods is found among incumbents in many markets. What cannot be moved. An executive does not choose who holds the shares. Nor does the executive choose their time horizon. What can be moved. What is explained, and over what horizon, is chosen. A company that explains only the quarter does not attract long-horizon shareholders. Shareholder composition is a given and also a consequence of explanation. A company with no language for Future Value is valued on financial value alone. That is not a limit of the market. It is a limit of disclosure (→ Vol. VIII, Ch. 075).

5.3 Decision by consensus

What cannot be moved. A behavioral pattern that prizes agreement does not change in a short period. Force a switch to unilateral decision and support disappears at the execution stage. What can be moved. Where the line falls between what requires consensus and what does not is drawn by the executive. Layers in the approval chain, monetary thresholds on decision rights, the number of sign-offs. All can be changed by board resolution. Consensus is not what makes decisions slow. The scope submitted to consensus is too wide. Scope is not an institution. It is a design.

5.4 The executive’s tenure and the composition of the board

What cannot be moved. The length of tenure is set by nomination mechanisms and convention. On a board composed largely of internal appointments, nomination easily becomes an extension of the internal hierarchy. This is sharpest in Japan, but boards dominated by internal promotees are common among long-established enterprises, and no single executive changes that alone. What can be moved. A design that refuses to align tenure with planning period is available. Align the medium-term plan with the term of office and the horizon always shrinks to that term. Further, what is handed to the next executive is the sitting executive’s to decide. What should be handed over is not a business. It is a question. Few companies have a design for passing an unfinished redefinition on while it is still unfinished.

5.5 Tolerance for failure

What cannot be moved. A company cannot control how society and the press react. A listed company is exposed to judgment at every disclosure. What can be moved. How failure is handled internally is management’s discretion. Most companies have institutionalized the decision to start and not the criteria for stopping. Because there are no criteria for stopping, withdrawal becomes defeat. Because it becomes defeat, no one starts. A low tolerance for failure is, before it is a problem of culture, a missing procedure.

5.6 The criterion that divides

Set the five side by side and one criterion emerges. Does changing it require a change in the law, or does a board resolution suffice? Whatever a board resolution can settle is not a constraint. It is a choice. Every item listed above under “what can be moved” belongs to the second category. Changeable things have been filed under the unchangeable. The problem of the incumbent is not that too much is unchangeable. It is that classification.

5.7 The same test applies to the side that holds nothing

New entrants face few institutional constraints. But the list of things “we cannot do yet” is just as long. The same shape of criterion applies here. Does changing it require someone else’s agreement, or does this week’s line-drawing suffice? The typical case is whether to partner with a large enterprise or compete with it. This is not a matter of nerve. It follows from four questions. Can the counterparty build the layer you hold for itself? On which side does the accumulation remain? Does the counterparty’s decision cycle match the time your capital has left? Is the dependence concentrated in a single firm (→ Vol. X, Ch. 092; Vol. X, Ch. 093)? The answer is not one answer for the whole business. It may differ layer by layer. No outside agreement is needed to draw that line.

5.8 Prescriptions differ by scale

Among incumbents, the center of gravity of the constraint shifts with scale. In large enterprises it sits in the interests of the existing business and the dispersion of decision rights. Nothing moves because no one holds enough authority to move it. The prescription is redesigned authority and stated criteria for stopping. Fix a share of the budget for uses that are not extensions of the existing business. That it is fixed is the point. In small and midsize enterprises it sits in thin capital and a thin bench of people, while decisions are fast. Purpose often already exists as the owner’s own words; what is needed is not to create it but to put it outside the company (→ 100 Questions on Management in the Age of AI, #037). One reservation. If cash flow fails, the Continuity term — the fifth element of Future Value — goes to zero. The theory does not deny that survival comes first. This reservation applies equally to both.

6 Questions for the executive

Everything so far reduces to two lines. What the enterprise that holds nothing must redefine is not its technology. It is the Purpose that says what it exists for, the design of capture that says how much of the value it creates stays with it, and the time axis over which it accumulates. What the enterprise that holds much must redefine is neither its business nor its organization. It is its own judgment about what cannot be changed. The two are not symmetrical. But both point at the same place. What do you accumulate, and how much of it stays with you? The enterprise that holds nothing has not yet chosen where accumulation happens. The enterprise that holds much has not measured how what it holds can move. Constraints are real. Labor law, shareholder composition, and the maturity of capital are not illusions. But few companies have ever measured the outline of the constraint. Left unmeasured, it widens year by year. What can be imported from another company is not speed. It is order. We close with three questions. Each can be answered at your next executive meeting. Both addressees get the same questions; only the place where they land differs. Question 1 — Can you write your core without using the name of a technology or the name of a current business? If you can, set the sentence beside the proposals you declined and the customers you chose this past year, and check for contradiction. An incumbent should check which line of this year’s budget it appears on. If it appears nowhere, it is a slogan, not a philosophy. Whether Purpose is a criterion can be verified only through the proposals you rejected, never the ones you approved. Question 2 — Of the things now described as impossible here, how many require someone else’s agreement? Make the list. In an incumbent most items fall inside what a board or the president can settle by decision. In a new entrant most fall inside what this week’s line-drawing can change. The work of separating constraint from excuse begins with that list. Question 3 — What sets the length of your time? A new entrant should check whether the maturity of its capital matches the life of its technology. If not, decide whether to shorten the business’s time or change the kind of capital. An incumbent should check whether the planning period coincides with the term of office, and which question it will hand to the next executive. A Future Horizon exceeds a term of office only where that design exists. None of the three questions asks how to get big fast. Each asks which premise you will doubt, when, and by your own decision. AI cannot answer that. AI can test a premise. Which premise to doubt is decided by a human being. AI Optimizes. Humans Define. First Principle 6 states, Enterprise Exists to Redefine Itself. That sentence records no founding year, no scale, and no nationality. Conditions change the speed. They do not change the direction. The first thing the enterprise that holds nothing must redefine is its answer to what it is doing this for. Foundation models can be rented. Computing capacity can be rented. People, given time, can be gathered. One thing cannot be rented. The first thing the enterprise that holds much must redefine is its own range of motion. Capital, people, and trust are already in hand. One thing is missing: a record of having measured, for itself, how far it moves.

In brief

One question, two answers. For the enterprise that holds nothing and the enterprise that holds much, the problem sits on opposite sides.
What the enterprise that holds nothing must redefine is Purpose, the design of capture, and its time axis.
What the enterprise that holds much must redefine is its own judgment about what cannot be changed.
Changeable things have been filed under the unchangeable. The criterion that divides is the level of resolution required.

Key concepts

Enterprise Redefinition / Enterprise Redefinition Capability / Future Capital / Future Horizon / Future Value Chain

The chain of ideas

Recognize → Enterprise Redefinition Capability → Capital Allocation

→ Future Time → Enterprise Value

Related first principles

Principle 1 — Purpose Precedes Profit. Principle 3 — Capital Exists to Create Possibility. Principle 6 — Enterprise Exists to Redefine Itself. Principle 8 — Trust Compounds Faster Than Capital.

Related chapters

Vol. V, Ch. 044 “What Is the Enterprise Redefinition Maturity Model (ERMM)?” — the instrument for diagnosing your own level across five dimensions
Vol. VI, Ch. 057 “How to Carry Out Enterprise Redefinition” — the seven stages turned into working procedure
Vol. VIII, Ch. 072 “What Is the Enterprise Value of a Startup?” — where the value of the founding years comes from
Vol. VI, Ch. 052 “What Does It Mean to Redefine the Executive?” — how not to mistake approval for management

Papers and companion volumes

Kadowaki, N. (2026). Future Value Theory: A Management Framework for Enterprise, Capital, and Society in the Age of AI. VURA Working Paper. SSRN: https://ssrn.com/abstract=7120980 / Zenodo: https://doi.org/10.5281/zenodo. 21255662
Kadowaki, N. (2026). Enterprise Redefinition: Toward an Enterprise Evolution Theory for the Age of AI. VURA Working Paper. (Published on Zenodo; under review at SSRN)
100 Questions on Management in the Age of AI, #058 “In the Age of AI, Do Startups Have the Advantage?” / #090 “Can Japanese Enterprises Win in the Age of AI?” / #037 “Why Small and Midsize Enterprises Can Win with AI”

Chapter 099 What Should New Entrants and Incumbents Redefine?