Chapter 084　What Did Apple Redefine?

Apple has been described as a company that makes products. Lay out its disclosures, however, and a different outline appears. Devices still carry the revenue. The structure of gross profit and the relationship with the customer, meanwhile, cannot be explained by the sale of devices. This chapter asks two questions. What did Apple redefine? And why is this company not always the fastest at redefinition? Building deeply and changing quickly are frequently incompatible. We treat that tension without celebrating it. Every figure carries its source and its date. We do not assert what was decided inside the company.

1 The question — what makes this company worth

asking about Begin with what can be confirmed. Apple’s revenue for fiscal 2025 (ended September 27, 2025) was $416.161 billion. Within it, iPhone was $209.586 billion and Services was $109.158 billion. Mac was $33.708 billion, iPad was $28.023 billion, and Wearables, Home and Accessories was $35.686 billion. These come from the fourth-quarter results disclosed on October 30, 2025. Services is about 26 percent of revenue. That percentage is our calculation from the disclosed values. On those figures alone, this still looks like a device company. Look at gross profit and the view changes. By the same disclosure, the gross margin for fiscal 2025 was 36.8% on Products and 75.4% on Services. Compute gross profit from the disclosed revenue and cost of sales, and Services accounts for a little over four-tenths of company-wide gross profit. A quarter of revenue produces four-tenths of the gross profit. That asymmetry is the most compact statement of what kind of enterprise Apple is. The scale keeps being updated. Revenue in the first quarter of fiscal 2026 (ended December 27, 2025) was $143.756 billion, of which Services was $30.013 billion (source: Apple Form 10-Q, filed with the SEC). Gross margin in that quarter was 40.6% on Products and 76.5% on Services. In a January 2026 announcement, the company said its active installed base had passed 2.5 billion devices. In the results announced on July 30, 2026, third-quarter revenue (for the quarter ended June 27, 2026) was reported at $109.4 billion, up 16 percent year on year. On the same day’s call, the company said paid subscriptions had passed 1.5 billion. The question follows from there. An enterprise with 2.5 billion active devices and 1.5 billion paid subscriptions is selling what, exactly? Devices? Or a continuing relationship for which the device is the entrance? Financially the two can produce similar numbers. As management they are entirely different undertakings. A second question sits on top of it. Against AI, the fastest-moving change of all, Apple is plainly not in front. In the development of foundation models and in the implementation of conversational assistants, the company arrived later than others. Its business has not shrunk. That fact puts a direct reservation against the argument about speed that we have been making since Vol. I. Changing fast is not by itself a value. Building deeply is not by itself a value. Which one, in which dimension — that choice is management. The Apple case is useful not as a success story, but because it is a rare instance in which the consequences of that choice can be observed over a long period.

2 Conventional answers and their limits

Three explanations of Apple circulate widely. Each is partly right. None is sufficient on its own. The first answer: “Apple is a brand company” The most frequently heard account. Consistency of design and experience produces a strong brand, and the brand supports a price advantage. The account explains what happens in the store well enough. Brand does not explain enormous investment in in-house silicon. If a brand can protect price, you can simply buy the components. The company announced the Mac transition to Apple silicon in June 2020, and in February 2025 it introduced its own modem, C1, in the iPhone 16e. In July 2026 it announced an expansion of its multi-year agreement with Broadcom to more than $30 billion. This is not brand-maintenance spending. It is spending to take design authority over cost, performance, and features back from component suppliers. The brand explanation cannot account for that allocation of capital. The second answer: “Apple earns by locking in an ecosystem” The second account looks at switching costs. Device, OS, apps, payments, and cloud interlock, and the customer finds it hard to leave. The account explains high retention well. Lock-in does not explain why customers come in to begin with. Blocking the exit and building the entrance are different designs. And there is a more important point. The premise of lock-in is itself being cut back by regulation. On July 8, 2026, the EU General Court dismissed Apple’s challenge to its designation as a gatekeeper. The obligation to open the App Store and iOS to competitors was upheld. In the United States, in connection with the Epic Games litigation, the Supreme Court was reported on May 6, 2026 to have declined Apple’s request for a stay. Settlement discussions between the company and the Department of Justice in its antitrust case were also reported as of July 2026. An account that treats lock-in as the body of competitive advantage runs out of road when that body is legally shrinking. The third answer: “Apple wins through vertical integration” The third account comes closest to the center. Hardware, OS, silicon, and services are designed by one company, so an experience emerges that others cannot copy. This looks like the model case treated in 100 Questions on Management in the Age of AI, #034, on the company that cannot be copied. Vertical integration, however, is not an unconditional advantage. Integration carries two costs. First, integration is fixed cost. The more territory you design yourself, the more you have to discard when you change direction. Second, integration slows some territories. A design philosophy that ships after the whole is coherent is slower than one that ships parts first. These costs surfaced in the AI phase. In March 2025 the company disclosed that delivery of a more personalized Siri would be delayed. Then, on January 12, 2026, a joint statement announced the adoption of Google’s Gemini as the foundation for Siri. Reporting put the agreement at roughly $1 billion a year; neither company has disclosed the amount. A vertically integrated company took an external model at its most strategic layer. All three conventional answers portray Apple as a company that is doing well. That is not what we want to see. We want to see what it let go, what it protected, and in which dimension the judgment was made.

3 What was redefined — read across the five dimensions

The five dimensions of Enterprise Redefinition are fixed in the order Purpose, Business, Organization, Capital, and Leadership. We read them in that order. One note first. The five dimensions do not change at the same frequency. In Apple’s case the spread is extreme.

3.1 Purpose — the core held, and its expression moved

What can be read from public information is that the company’s Core Purpose has been stable over a long period. The core is to finish technology into a tool for an individual to use. That core has not been rephrased even as product categories were replaced. What moved is the expression. Entering the 2020s, the company began to speak of privacy not only as a property of a product but as a premise of the business. This has been carried into the design of AI features. At WWDC26 on June 8, 2026, the company introduced the next generation of Apple Intelligence and Siri AI. Processing is described as split between an on-device model and Private Cloud Compute. A design document for Private Cloud Compute is published on the company’s security research blog. This needs to be read carefully. Raising privacy is a statement of ethics and, at the same time, a competitive position. We treat it as an instance of Purpose in its expression connecting to business design. We do not assert motive.

3.2 Business — from a single sale to a continuing relationship

This is the dimension that moved most. In a business that sells a device outright, the handover of value ends at the moment of purchase. This company has thickened the layer of revenue that begins after purchase. Services revenue was $109.158 billion in fiscal 2025 and $30.013 billion in the first quarter of fiscal 2026. Paid subscriptions were said to have passed

1.5 billion as of the quarter ended June 2026.

It is important not to mistake the managerial meaning here. Adding subscriptions is not the redefinition. The redefinition is that the length of time with the customer became the unit of value. The device becomes the entrance to the relationship, and the quality of the relationship supports the replacement of the device. Once that circuit is turning, the company is in a different business. The contents of Services, however, include parts that depend on other companies. Agreements over default settings in search are one example. In September 2025 it was reported that a remedies ruling by a US federal district court permitted such agreements to continue on conditions. Had the ruling gone the other way, the composition of Services might have changed. A business that sells a continuing relationship also stands on judicial decisions made elsewhere.

3.3 Organization — the boundary of design was redrawn

The organizational redefinition here is not a rearrangement of departments. It is the movement of the boundary of how much the company designs itself. The June 2020 announcement of the Apple silicon transition is the clearest example. Taking processor design in-house means installing semiconductor design as a permanent capability in the organization. The C1 modem introduced in February 2025 moved the boundary further. The agreement with Broadcom of more than $30 billion in July 2026 is a move to secure production of wireless components inside the United States. As the inside of the boundary widens, the organization becomes stronger and heavier at once. The weight appears as cost at the moment of change. There is a second effect, easily missed. The wider the inside of the boundary, the narrower the channel through which outside technical progress enters. When building it yourself is the premise, the motive to evaluate other companies’ results weakens. The benefit of integration is coherence of design. The cost of integration is distance from external learning. That distance can be read as having surfaced in the AI phase.

3.4 Capital — away from the past, toward design authority

In the Capital dimension, allocation moved from component procurement to design capability. In August 2025 the company announced that it was raising its US investment commitment to $600 billion over four years. Through 2026, announcements have continued, including additional manufacturing production of the Mac mini. partners and US Here the canon’s diagnostic question bites. Are resources allocated toward Future Value rather than historical success? The company’s capital allocation goes not only to producing more of existing products but to design authority over components and manufacturing. At the newest layer of all, foundation models, a structure of dependence on external agreements has emerged. The center of gravity of capital allocation and the center of gravity of technological change do not fully coincide.

3.5 Leadership — a succession is itself a statement of design

On April 20, 2026, the company announced that Tim Cook would become Executive Chairman and that John Ternus would become CEO. The appointment takes effect on September 1, 2026. Ternus joined in 2001 on the product design team and has served as senior vice president of hardware engineering since 2021. Where a successor comes from is also a statement about what management places at the center. Here again we avoid assertion. Choosing a successor from the hardware design lineage can be read as an intention to continue with integration. Whether that reading is right can only be tested over the next several years.

4 Structure — maturity, the value chain, and the

equations Put the facts onto the skeleton of the theory.

4.1 Where the company sits in the Enterprise Redefinition

Maturity Model The Enterprise Redefinition Maturity Model (ERMM) has five levels: Reactive Enterprise, Improvement Enterprise, Transformation Enterprise, Continuous Redefinition Enterprise, and Future Value Enterprise. Apple cannot be placed at a single level. As the source paper states explicitly, progression is not necessarily linear. Within one enterprise, the level splits by dimension. Within what public information shows, the Business and Capital dimensions display behavior that treats change as normal work rather than as a project. Bringing silicon in-house was not a single transformation but a continuous redesign running over a decade. That behavior is close to the description of a Continuous Redefinition Enterprise. The response to the new layer of AI does not look like the same level. Delayed features were disclosed, and the sequence ended in the adoption of an external model. That is periodic rather than continuous. Taken on its own, it is close to the description of a Transformation Enterprise. What matters is that this assessment is not a ranking. The Enterprise Redefinition Maturity Model evaluates organizational coherence rather than isolated excellence, and maturity is assessed across all five dimensions in balance. Exceptional capability in one dimension does not raise the whole. When levels split by dimension, an enterprise tends to run its weak dimension on the logic of its strong one. Set the release date for AI on a hardware standard of completeness and you will be late. Design hardware at the iteration speed of AI and quality will break. What Apple faces can be read as this problem of translation. And, as the canon repeats, Level 5 is not to be treated as a target to be reached as fast as possible. The appropriate level differs by industry and environment. No outsider can specify which level Apple should aim for.

4.2 Where in the Future Value Chain the value was created

The order of the Future Value Chain is fixed. Purpose → Learning → Redefinition → Creation → Enterprise Value Where in that chain was Apple’s value created? What can be read is that value is concentrated between Redefinition and Creation — that is, in the integration of design. Decide again what the company will design itself, then make that decision real as a product. That interval is where this company is strongest. The weak interval is Learning. In the AI phase, the company’s learning surfaced later than others’. When learning is late, redefinition starts late. When redefinition is late, creation is late. The chain clogs in order. And Enterprise Value comes last. That the company’s financial results are strong is the result of redefinitions completed in the past. Today’s numbers do not prove today’s capacity to redefine. Drop that distinction and a case study becomes a tribute.

4.3 Reading the case through two equations

The capability to create Future Value is expressed by the following formula. FVCC = Purpose × Learning × Redefinition × AI Integration × Ecosystem × Capital Allocation × Trust This is multiplication, not addition. If any one term approaches zero, the whole shrinks however large the others are. Applied to Apple, the Purpose, Redefinition, Ecosystem, Capital Allocation, and Trust terms have historically been high. Learning and AI Integration, at least over the last several years, have not been at the level of the others. Because the relationship is multiplicative, those two terms hold the whole down. Within what public information shows, the decision to adopt Gemini can be read as a choice to supply those two terms from outside. Take the Trust term on its own. Value = Purpose × Trust × Capability × Time A policy built around privacy works on the Trust term of this equation. And, as the eighth of the First Principles states, Trust Compounds Faster Than Capital. Trust grows faster than capital. Compounding, however, also runs in reverse. An enterprise that keeps its design closed on the grounds of trust will be asked by regulators to distinguish closure for safety from closure that excludes competition. The EU decision makes exactly that distinction the issue.

5 What it looks like in practice — the turning points, and

what was let go Set the events in sequence and the character of the judgments becomes visible. June 2020, the Mac transition to Apple silicon is announced. February 2024, cancellation of the automobile project is widely reported. February 2025, the iPhone 16e with the C1 modem is announced. March 2025, a delay to more personalized Siri features is disclosed. August 2025, the US investment commitment is raised to $600 billion over four years. September 2025, continuation of the search default agreement on conditions is reported. January 12, 2026, the adoption of Google’s Gemini is announced in a joint statement. April 20, 2026, the CEO succession is announced. June 8, 2026, Siri AI is introduced at WWDC26. July 8, 2026, the EU General Court dismisses the challenge to the gatekeeper designation, and on the same day the expanded Broadcom agreement is announced. Three things can be read from that sequence. First, the company chooses order rather than speed. Bringing silicon in-house proceeded over more than five years. The automobile was cancelled after nearly a decade of development. Time goes into deciding whether to ship at all, rather than into shipping quickly. Second, what it let go is large. It let go of dependence on Intel processors. It let go of part of its external modem procurement. It let go of the automobile as a new business altogether. And in 2026 it suspended, for foundation models, part of the principle that the most important layer is always built in-house. That fourth item is the most important fact in this chapter. An enterprise whose creed is vertical integration partnered externally at the core of AI. This can be read as a record of defeat, and it can be read as a record of swapping the means in order to protect the Purpose. We hold that public information alone cannot decide between the two. Third, slowness has consequences in both directions. Being late, the company avoided much of the quality trouble that arose in the early period of AI features. Being late, it also gave up part of the initiative in the design of AI experience. Which way that balance falls cannot be judged until the results of the next several years are in. About the future, nothing is available but a reservation.

5.1 Dependence and risk, written as part of the case

To avoid tribute, we set out the risks explicitly. First, concentration of revenue. iPhone revenue in fiscal 2025 was $209.586 billion, half of company-wide revenue. Concentration in a single product is fragile against a change in demand. Second, external dependence. Part of Services depends on agreements with other companies, and a structure has emerged in which the core of AI depends on an agreement over an external model. A vertically integrated enterprise is most externally dependent at its newest layer. Third, regulation. The judicial decision on the EU’s DMA, the Epic-related litigation in the United States, and the Department of Justice antitrust case can all affect the economic terms of the App Store. The margin of Services stands on those terms. Fourth, the leadership transition. The CEO changes on September 1, 2026. The transition reads as having been prepared deliberately, but the consequences of a succession appear over several years. Here too we reserve judgment. None of these is a reason the company falls tomorrow. They are, however, reasons that today’s financial results must not be read as evidence of the capacity to redefine.

6 What transfers, and questions for the executive

Many incumbents have made vertical integration their strength. The Apple case offers that group four implications. First, integration is a means, not an end. The line marking what the company designs itself should be redrawn every year. Apple moved the line inward for processors and modems, and outward for foundation models. An enterprise that fixes the line turns integration into a creed, and the creed eventually becomes a liability. The question is not “in-house or not.” It is “can the Purpose be realized without designing this layer ourselves?” Second, estimate the cost of slowness dimension by dimension. Slowness spent building quality accumulates as trust. Slowness caused by late learning accumulates nothing. This is where incumbents frequently confuse two things. Care and indecision are not the same. The first becomes an asset, and the second becomes a loss. Third, remeasure the unit of the business in time. Can you define the business by the number of years the relationship with the customer continues, rather than by the value of a single sale? The question bites hardest in manufacturing. Apple did not discard the device; it changed what the device means. What an incumbent needs is not to discard products. It is to change what products mean. Fourth, build regulation into design as a premise of the business. A design that depends on the economics of lock-in is fragile against changes in law. Can you move to a design that maintains the relationship through reasons to keep being chosen, rather than through enclosure? This is a structural question before it is an ethical one. Three questions to close. Each can be answered at your next executive meeting. Question 1 — Among the areas you design in-house, which are indispensable to realizing your Purpose? In-house work that is not indispensable is inheritance. Inheritance looks like an asset and binds judgment. Hold a meeting to redraw the line at least once a year. Question 2 — Is your slowness the slowness of building, or the slowness of not deciding? Separate the two things spoken of with the same word. The first may continue. The second must be fixed within this fiscal year. An organization that cannot tell them apart preserves all of its slowness as a virtue. Question 3 — Today’s strong results are the outcome of a redefinition made how many years ago? If the answer is ten years ago, you must be able to say what you are redefining now. When you cannot say it, the enterprise is consuming a past success. As the sixth of the First Principles states, Enterprise Exists to Redefine Itself. An enterprise exists in order to redefine itself. Apple moved from a company that sells products to a company that sells an integrated experience and a relationship. The move was not fast. That it was not fast is both a strength and a weakness. What we should take from this case is neither a celebration of integration nor a celebration of speed. It is the question of whether you are deciding for yourself which dimension moves at which speed.

In brief

What Apple redefined is the relationship with the customer: from a single sale to something measured by length of time.
On what public information shows, Business and Capital read as a Continuous Redefinition Enterprise, while the AI layer is closer to a Transformation Enterprise.
Value was created between Redefinition and Creation, in the interval called the integration of design.
If the lag in Learning persists, integration as a strength turns directly into external dependence.

Key concepts

Enterprise Redefinition / the Enterprise Redefinition Maturity Model / Future Value Chain / Future Value Creation Capability / the Ownership-to-Usage Pattern (→ Vol. VI, Ch. 059)

The chain of ideas

Purpose → Redefinition → Creation → Trust → Enterprise Value

Related first principles

Principle 4 — AI Optimizes. Humans Define. Principle 6 — Enterprise Exists to Redefine Itself. Principle 8 — Trust Compounds Faster Than Capital.

Related chapters

Vol. VI, Ch. 059 “Cases of Enterprise Redefinition” — the definition of the Ownership-to-Usage Pattern and the conditions for its success
Vol. V, Ch. 050 “What Does It Mean to Redefine Customer Value?” — putting the length of the relationship in place as the unit of value
Vol. V, Ch. 044 “What Is the Enterprise Redefinition Maturity Model (ERMM)?” — how to read levels that split by dimension
Vol. VII, Ch. 068 “Does Trust Become Enterprise Value?” — the relationship between closed design and trust

Papers and companion volumes

Kadowaki, N. (2026a). Future Value Theory: A Management Framework for Enterprise, Capital, and Society in the Age of AI. VURA Working Paper Series. SSRN: https://ssrn.com/abstract=7120980 / Zenodo: https://doi.org/10.5281/zenodo. 21255662
Kadowaki, N. (2026b). Enterprise Redefinition: Toward an Enterprise Evolution Theory for the Age of AI. VURA Working Paper Series. (Published on Zenodo; under review at SSRN)
100 Questions on Management in the Age of AI, #034 “What Is a Company That Cannot Be Copied in the Age of AI?”

Chapter 084 What Did Apple Redefine?