← Back to Companies

Technology

Apple

Customer-Favorable Central finding Dependency, not price Scope Apple consumer ecosystem, US market Methodology CHI/CVI v2.0 Last revised August 2026

The more Apple gives you, the harder Apple becomes to replace.

Apple's Customer Hostility Index finding is not primarily about expensive hardware. It is about dependency. Apple creates extraordinary value by making its products work better together. But every additional device, service, account and pool of data can also increase the number of relationships a customer must reconstruct to leave. That creates two simultaneous forces: an Integration Gradient of increasing customer value, and a Dependency Gradient of increasing customer exit burden. They are not separate systems. They are the same architecture, measured from two ends.

CHI68/100Substantial dependency
CVI84/100Exceptional
CFS+16Customer-Favorable
Integration Gradient More Apple → more value Each additional product raises interoperability, synchronization and convenience.
Dependency Gradient More Apple → more to rebuild Each additional product adds a relationship that must be reconstructed to leave.

CFS = CVI − CHI (84 − 68 = +16). Assessed under CHI/CVI Methodology v2.0: dimensions score the underlying customer harm, patterns describe the behavior that produced it. Revised August 2026 following the 28 July 2026 US launch of Apple Upgrade, Apple’s consumer hardware leasing program: CHI 66 → 68, CVI unchanged at 84, CFS +18 → +16. The classification band is unchanged. CHI 68, CVI 84 and CFS +16 are the settled analytical values and are not presented as a range. Overall classification: Customer-favorable — with substantial dependency. Methodology →

Finding Heatmapⓘ
Accumulated ecosystem dependency
Dependency Monetization (storage)
Residual exit resistance
Repair control (genuine-parts economics)
Rentalization (hardware leasing)
Controlled interfaces (historical)
First-party behavioral steering
First-party advertising & consent
Planned obsolescence (general)*
Not established
Strongly supported
* Planned obsolescence as a general Apple strategy is shown at minimum weight because it did not survive falsification testing. It appears here so the reader can see what was tested and rejected, not because the page asserts it.

The defining finding

Nothing traps the customer. The accumulation does.

No single Apple relationship is a trap. One phone can be replaced. One account can be closed. The finding is cumulative: hardware, identity, family, purchased media, synchronized data and household infrastructure each attach separately, and each must be reconstructed separately. Strongly supported Apple does not have to make leaving impossible. It only has to make leaving a project. Analytical inference

Can leave ≠ easy to leave. That gap is what CHI is measuring.

CHI dimension scores (v2.0 families, 0–5 severity)
Customer Restriction4.5/5
Behavioral Manipulation3.25/5
Information & Privacy3.25/5
Revenue Extraction3.25/5
Trust & Transparency2.75/5

The central tension · Primary exhibit

The Apple Paradox.

Read both columns as one sentence. They describe the same design decision — that Apple products should work better together — measured once as benefit and once as burden.

Integration Gradient

More Apple

  • More interoperability
  • More synchronization
  • More convenience
  • More functionality
  • More incremental value

Dependency Gradient

More Apple

  • More accumulated data
  • More connected hardware
  • More identity relationships
  • More family relationships
  • More infrastructure to reconstruct
SAME ECOSYSTEM. TWO EFFECTS. The architecture that creates Apple's greatest customer value also creates its greatest customer dependency.

Primary finding · Customer / Platform Lock-In

Apple does not prevent customers from leaving. It increases what leaving costs.

Customer / Platform Lock-In — 4.5 / 5 · HIGH

Apple's strongest current hostility finding is accumulated ecosystem dependency. Apple does not necessarily prevent customers from leaving. It increases the number of things customers must replace to leave. Strongly supported

An iPhone alone creates relatively limited dependency. An iPhone connected to a Mac, Watch, AirPods, iCloud Photos, Passwords, Health, Find My, Family Sharing, Home and years of accumulated data creates something substantially different.

No individual relationship necessarily traps the customer. The accumulation creates the burden.

Why this scores where it does

CHI scores the customer harm, not the intent. A dependency finding does not require a finding that Apple engineered the dependency to trap anyone: the burden on the customer is the same whether it is the deliberate product of lock-in strategy or the incidental product of unusually good integration. Analytical inference

It also does not require a finding that the customer is worse off. Most of these relationships are things the customer actively wanted and continues to benefit from — which is precisely why they accumulate, and precisely why the exit burden grows quietly rather than visibly.

Dependency is scored as a measured customer cost, not as an accusation of intent.

The Dependency Gradient · How the burden accumulates

The same company. Three very different exits.

Nothing changes about Apple between these three stages. What changes is how many relationships the customer has accumulated — and therefore how many must be reconstructed elsewhere.

Light dependency iPhone

Relatively straightforward substitution. One device, one account, one set of data to move.

Moderate dependency iPhone + Mac + Watch + AirPods + iCloud

Multiple hardware, data and workflow relationships must be replaced. Some have no direct equivalent outside the ecosystem, and several stop working the moment one piece is substituted.

Deep dependency Apple household + Family Sharing + Photos + Health + Passwords + Home + purchased apps/media + years of data

Departure becomes a multi-system migration project — and one that is no longer a single customer's decision, because family members, shared purchases and household infrastructure are attached to it.

CAN LEAVE ≠ EASY TO LEAVE

This is a structural illustration of accumulation, not a measured migration-time metric. The finding is the shape of the curve, not a number of hours.

CHI dimension breakdown · v2.0

Where the 68 comes from.

Under v2.0 the five dimensions score the underlying customer harm and the patterns describe the behavior that produced it. Apple's score is concentrated in one dimension; the remaining four are moderate, and one is the lowest finding on the page. Two dimensions moved in the August 2026 revision; three did not.

Customer Restriction — 4.5 / 5 Primary mechanism: ecosystem dependency, residual exit friction and repair control. This is the dimension that carries the Apple assessment, and it is the only one scored in the high band.
Behavioral Manipulation — 3.25 / 5 First-party commercial prompting and limited consent/choice-architecture findings. Moderate rather than severe: the mechanisms are real but are not the social-platform engagement machinery scored elsewhere in the index.
Information & Privacy — 3.25 / 5 Specific first-party advertising and consent concerns despite unusually strong privacy architecture. Both halves are held at once: the architecture is genuine, and the first-party carve-outs inside it are the finding.
Revenue Extraction — 3.25 / 5 (was 3.0) Primarily dependency monetization rather than ordinary hardware Price Creep. The mechanism scored here is the monetization of accumulated relationships, not the headline price of the hardware. Raised in August 2026 for the addition of Rentalization as a second, independently supported extraction mechanism: consumer hardware leasing through Apple Upgrade.
Trust & Transparency — 2.75 / 5 (was 2.5) Historical control-preserving behavior exists, but broad minimum-compliance theories did not survive. Still the lowest of the five, and it is lower because claims were rejected, not because they were never tested. Raised in August 2026 for verified quantification gaps in Apple Upgrade: the damage fee, the month-to-month payment increase and the buyout are each described in Apple’s terms and none is sized.
Reading these numbers Dimension scores are severity on a 0–5 scale and are not additive into the CHI total by simple summation; the v2.0 weightings differ by dimension (25 / 25 / 20 / 15 / 15). August 2026 delta: Revenue Extraction 3.0 → 3.25 and Trust & Transparency 2.5 → 2.75, both attributable to Apple Upgrade; Customer Restriction, Behavioral Manipulation and Information & Privacy are unchanged. CHI moves 66 → 68. No additional numeric family scores are published for Apple beyond the five above. Methodology →

iCloud · Dependency Monetization

The price is not the finding. The ratchet is.

The mechanism runs in one direction, and every turn of it makes the next turn easier.

Use
Accumulate data
Exhaust free storage
Subscribe
Accumulate more
Deeper dependency

Pattern: Dependency Monetization

The finding is not that iCloud is outrageously priced. It isn't. The concern is that a genuinely useful service becomes increasingly difficult to unwind as the customer's accumulated data grows. Strongly supported

That is a different mechanism from Price Creep, and it is scored differently. Price Creep is a rising charge for a stable relationship. Dependency Monetization is a stable charge for a relationship that becomes progressively harder to end — the customer's own accumulated data is what raises the switching cost, and it raises it whether or not the price ever moves. Analytical inference

A service can be fairly priced and still be difficult to leave. Those are separate questions, and CHI asks the second one.

Apple Upgrade · Rentalization · New, August 2026

Ownership becomes the option you have to ask for.

On 28 July 2026 Apple launched Apple Upgrade in the United States — a consumer lease provided by Klarna Inc. covering iPhone, Apple Watch, iPad and Mac. In the same announcement Apple said it would “no longer offer the iPhone Upgrade Program and iPhone Payments in the United States,” retiring the instalment route that ended in ownership. Apple Card Monthly Installments, which does end in ownership, is unaffected. Fact

Lease
Lower monthly payment
Term ends
Return device
Lease again
No asset accumulates
Apple’s four published worked examples. Apple publishes the purchase price and the monthly payment; the total, the share and the buyout are CHI arithmetic on Apple’s own figures. Longer of the two available terms shown. Excludes tax and any trade-in credit. Source: Apple Newsroom, 28 July 2026.
DevicePurchase priceTermMonthlyTotal of paymentsShare of price paidBuyout to ownTotal if you own
iPhone 17 Pro 256GB$1,09924 mo$31.99$767.7670%$331.24$1,099
Apple Watch Series 11 42mm$39924 mo$11.99$287.7672%$111.24$399
iPad Pro 11-inch 256GB$1,19936 mo$24.99$899.6475%$299.36$1,199
MacBook Pro 14-inch 16GB$1,99936 mo$38.99$1,403.6470%$595.36$1,999

Scroll the table horizontally →

Two things are true in the same table. A customer who buys out pays exactly the purchase price — there is no finance charge, and no route ending in ownership costs more than buying outright. A customer who returns has paid 70 to 75 per cent of the price of a device and owns nothing.

The finding

Ownership stops being the default and becomes a lump sum. Apple states the purchase option as a formula — purchase price less payments already made — and does not perform that arithmetic on a real product anywhere on its consumer pages; the single worked example uses a round $1,000 hypothetical. No total-of-payments figure appears on any Apple Upgrade page, and the product grid shows a monthly figure without the purchase price beside it. Fact

Returning is not free. The lessee gives up a residual Apple prices at 25 to 36 per cent of the purchase price. Independent analysis puts the two-year market value of a returned iPhone 17 Pro at roughly $410 through Apple Trade In and $530–$550 in private resale, against the $331 buyout forgone. Analytical inference

The marketed exit is another lease. Upgrading early costs the remaining payments in full, so the only fee-free upgrade point is the end of the term — which is where the next lease begins. Fact

A lower monthly payment and no asset at the end are the same fact, read from two ends.

The strongest counterargument — and it is strong

There is no finance charge. Because the buyout is the purchase price less payments made, every route ending in ownership costs sticker plus tax. Apple states it plainly: “If you return your device in good working condition, you will not pay more than the full price of the device.” Fact

On Mac and iPad the lease is a better zero-per-cent offer than Apple’s own. Apple Card Monthly Installments runs 12 months on a Mac; Apple Upgrade runs 36. A $1,999 MacBook Pro is $166.58 a month for a year under ACMI, or $38.99 a month for three years plus a $595.36 buyout under the lease — $1,999 either way. Fact

The terms are unusually light. Soft credit inquiry only, no down payment, nothing due at signing, no late fees, no origination or monthly fees; Apple has stated there is no device lockout for missed payments. Doing nothing at the end of the term does not repossess the device — after a six-month extension the customer is charged the buyout and keeps it. Fact

Nothing here is priced above sticker. What CHI asks is what the customer holds when the payments stop.

Who owns it Klarna Inc. is the lease provider; the customer is a lessee. Apple answers the question directly — “Do I own my device at the end of my lease term? No, you do not.” The disclosure exists, in Apple’s own words. It sits in a collapsed FAQ and a legal footnote. Fact
What is not sized The damage fee has no published schedule, cap, or definition of “good working condition.” The month-to-month extension “may increase” the payment, with no rate or cap given. Both are charges the customer cannot compute before signing. Fact
What the trade-in becomes A trade-in lowers the monthly payment across the initial term rather than creating equity, cannot be applied to a later upgrade, and can be revalued — raising the payment — if the traded device fails inspection. Fact
What was withdrawn The retired iPhone Upgrade Program included AppleCare+ and ended in ownership; Apple Upgrade bills AppleCare separately. Two independent analyses find the lease costs more per month than the retired programme once AppleCare+ with Theft and Loss is added back. Strongly supported

Classified: Rentalization — CONFIRMED, moderate.

What is not being claimed. Apple hardware remains fully purchasable, outright or on interest-free instalments, and Apple Upgrade is one of four routes Apple offers. The pattern is not scored because leasing exists. It is scored because Apple withdrew the ownership-terminating financing programme in the same announcement that introduced the lease, and because the default outcome of the lease — return, or lease again — leaves the customer with no asset after paying 70 to 75 per cent of the device’s price. Fact

Scored in Revenue Extraction as a second, independently supported extraction mechanism alongside Dependency Monetization; that dimension moves 3.0 → 3.25. The quantification gaps are a distinct harm and are scored once, in Trust & Transparency, which moves 2.5 → 2.75. The carrier condition attached to iPhone leases — a postpaid AT&T, T-Mobile or Verizon plan, with prepaid and MVNO customers excluded — is recorded as corroborating Customer Restriction but is not rescored there: that dimension is already carried by accumulated ecosystem dependency, and one additional condition in one market does not move it. Anti-double-counting rule →

Context, not a finding. Apple Upgrade launched four weeks after Apple raised Mac, iPad, Apple TV, HomePod and Vision Pro prices by up to $300 — roughly 13 to 25 per cent — on 26 June 2026, citing AI-driven memory demand. CHI excludes ordinary price movement from hostility scoring, and the shortage is industry-wide and documented. The sequence is recorded because the programme presents those prices as a monthly figure. No claim is made that the programme was created to disguise them. Not established

Hardware transitions · Falsification discipline

When design changes create customer cost.

Each of these is the site of a widely repeated accusation. In every case the customer cost is supported and the motive claim is not. The distinction is the point of this section: a real cost imposed on customers does not license an unevidenced claim about why.

Headphone Jack Supported: customer transition cost.
Unsupported: removed specifically to force AirPods purchases. Not established
Charger Removal Supported: customer cost-shifting.
Unsupported: a simple accessory-revenue conspiracy. Not established
Lightning / MFi Strongly supported: historical interface control.
Current burden: substantially reduced following USB-C. Strongly supported
Repair Supported: Apple control over genuine-parts economics.
Unsupported: repair prices manipulated to sell AppleCare+. Not established

Lightning is retained in the record as a historical finding and is not converted into a present-tense current claim. Where a burden was materially removed, the score reflects its removal.

The evidentiary spine · Second priority exhibit

Supported vs Not Established.

The right-hand column is not a disclaimer. It is the reason the left-hand column can be trusted.

SUPPORTED

Findings that survived falsification testing and are carried into the score.

  • Ecosystem dependency
  • Customer / Platform Lock-In
  • Dependency Monetization
  • Residual Exit Resistance
  • Repair Control
  • Rentalization — hardware leasingNew, August 2026. The ownership-terminating iPhone financing programme was retired in the same announcement.
  • Historical controlled-interface monetizationRetained as historical; current burden materially reduced.
  • Moderate first-party behavioral steering
  • Specific first-party privacy / consent concernsScored alongside, not instead of, Apple's privacy architecture.

NOT ESTABLISHED

Claims tested during the investigation that the evidence did not support. They are not asserted anywhere on this page.

  • Headphone jack removed specifically to force AirPods purchases
  • Repair pricing deliberately manipulated to sell AppleCare+
  • Apple Upgrade costs more than buying outrightThere is no finance charge; the buyout is the purchase price less payments made, so ownership costs sticker either way.
  • Apple captures the residual value of returned lease hardwareNeither Apple nor Klarna discloses who takes title to returns or where they go. Klarna’s CFO was asked directly on 18 August 2026 and did not allocate the risk.
  • Leasing was introduced to disguise the June 2026 price increasesThe sequence is real. No evidence of intent was found in either direction.
  • Broad current Apple hardware unreliability
  • Planned obsolescence as a general Apple strategy
  • Artificial Scarcity as a material Apple mechanism
  • Social-platform-style Infinite Engagement
  • "Apple privacy is fake"

The investigation was designed to identify hostility—not manufacture it.

Remediation

Apple has also reduced dependency.

Several of the burdens most often cited against Apple have been visibly reduced. CHI records that, because a score that only moves in one direction is not measuring a relationship — it is stating a conclusion.

Voluntary / standards-led

Adopted alongside open industry standards

Improved passkey portability — credential portability built on shared industry standards rather than an Apple-only format.
Qi2 — wireless charging on a common standard rather than a proprietary one.
Matter — smart-home interoperability across vendors, which directly reduces the household-infrastructure component of the Dependency Gradient. Fact

Each of these makes a specific dependency relationship reconstructable elsewhere.

Regulatory-pressure correlated

Improvements that coincide with regulatory pressure

Improved outbound migration and Transfer to Android — data-portability routes out of the ecosystem.
USB-C — the single largest reduction in the controlled-interface finding.
Repairability improvements and expanded parts support.
Broader interoperability across previously closed surfaces. Fact

Correlation with regulatory pressure is recorded. Causation is not asserted.

CHI scores the current customer relationship. A company does not remain permanently penalized for a burden it genuinely removes.

These two groups are presented separately because they are different kinds of evidence, not because one is sincere and the other is not. Several items appear in a regulatory context and would also have been commercially rational independently; this page does not attempt to resolve which motive was operative, and does not treat the two categories as interchangeable. Analytical inference

The value counterweight

Why CVI is still 84.

The customer value here is not a courtesy paragraph. It is the larger of the two numbers on this page, and it constrains every hostility finding above it.

Core Product Value4.25/5
Experience & Convenience4.25/5
Security, Privacy & Protection4.25/5
Longevity & Improvement4.0/5
Ecosystem / Incremental Value4.25/5

The audit that moved the number

Apple's provisional CVI was 92. An adversarial falsification audit reduced it to 84 after identifying material recent software, account, synchronization, security-remediation and ecosystem-delivery failures. Strongly supported

The value score was attacked as hard as the hostility score. It moved eight points.

Where the deductions came from

The hardware attack largely failed. The strongest deductions came from execution, not from the fundamental product architecture. Analytical inference

That distinction matters for reading the Ecosystem / Incremental Value score in particular: the same integration that produces the dependency finding above is genuinely delivering value here, and is scored as such.

A high CVI is not a defence against the CHI. It is the other half of the same measurement.

August 2026 re-test: CVI holds at 84.

Apple Upgrade is a payment structure, not a product capability, and does not map onto any CVI dimension: it adds a genuine zero-per-cent option with a longer term than Apple’s own instalment plan, and removes an instalment programme that bundled AppleCare+. Those roughly offset, and neither reaches the 0.25 granularity of a family score. The two open items the falsification audit named as capable of moving a family were both re-examined. The second ITC proceeding over Apple Watch blood oxygen terminated in Apple’s favour in April 2026, but the degraded implementation the audit priced is unchanged. The rebuilt Siri is shipping in the iOS 27 betas to qualified reviews, is not yet generally available, and is unavailable in the EU and China — the audit’s condition for restoring Core Product Value to 4.5 was “ships and performs,” and it is not yet met. No family moves. Analytical inference

Final balance

Two numbers, measured independently.

68Hostility
84Value
+16Customer Fairness

Apple remains customer-favorable.

But Apple's high value does not erase its hostility. And its hostility does not erase its value. That is why CHI and CVI are measured independently.

The margin narrowed in August 2026, from +18 to +16, and the classification band did not change. It is worth saying where the band ends: below +15 the relationship is classified as finely balanced rather than customer-favorable. Apple is two points above that line.

Explore the investigation

Nine research areas behind the score.

Each area below is summarized to its finding rather than reproduced in full. Where an area has a dedicated section on this page, the card links to it.

Dependency & Ecosystem Architecture → The Integration Gradient and the Dependency Gradient are the same architecture read from two ends. This is the organizing finding of the whole assessment and the reason Customer Restriction scores 4.5 / 5. iCloud & Storage Dependency → Dependency Monetization rather than Price Creep: a fairly priced, genuinely useful service whose difficulty of exit rises with the customer's own accumulated data. Migration & Exit Resistance → Residual exit resistance is SUPPORTED, and materially reduced by improved outbound migration and Transfer to Android. The remaining burden is accumulation-driven: it scales with the household, not with any single barrier. Apple Upgrade & Rentalization → New in August 2026. Apple’s US hardware lease shifts the default outcome of a Mac, iPad, iPhone or Watch purchase from ownership to return, with no finance charge and no premium over sticker for anyone who buys out. Rentalization is CONFIRMED at moderate severity; Revenue Extraction moves 3.0 → 3.25 and Trust & Transparency 2.5 → 2.75. Hardware Transitions & Complementary Purchases → Headphone jack, charger removal and Lightning / MFi. Customer transition cost and historical interface control are supported; the motive claims attached to them are not, and USB-C substantially reduced the current burden.
Repair & AppleCare+ Apple control over genuine-parts economics is SUPPORTED and contributes to Customer Restriction. The claim that repair pricing is manipulated to drive AppleCare+ attachment was tested and is NOT ESTABLISHED. Repairability improvements and expanded parts support are recorded as remediation.
Behavioral Manipulation 3.25 / 5. First-party commercial prompting and limited consent/choice-architecture findings. Social-platform-style Infinite Engagement was tested against Apple's first-party surfaces and NOT ESTABLISHED; the moderate score reflects prompting and default placement, not attention capture.
Information & Privacy 3.25 / 5. Specific first-party advertising and consent concerns sit inside an unusually strong privacy architecture. "Apple privacy is fake" was tested and REJECTED; the finding is narrower and concerns first-party carve-outs and consent presentation, not the architecture itself.
CVI Falsification Audit Provisional 92 → final 84. An adversarial review identified material recent software, account, synchronization, security-remediation and ecosystem-delivery failures. The hardware attack largely failed; the deductions are execution findings, not architecture findings.

Associated patterns · Final state

The patterns materially relevant to this verdict.

Primary

Secondary

Dependency Monetization CONFIRMED — a useful, reasonably priced service that becomes progressively harder to unwind as the customer's accumulated data grows. Scored distinctly from Price Creep.
Rentalization → CONFIRMED, moderate — new in August 2026. Apple Upgrade leases iPhone, Watch, iPad and Mac in the US through Klarna, and the ownership-terminating iPhone financing programme was retired in the same announcement. Scored in Revenue Extraction; the claim that leasing costs more than buying is NOT ESTABLISHED and is not asserted. Exit Resistance → SUPPORTED, residual. Materially reduced by improved outbound migration and Transfer to Android; the remaining burden scales with accumulation rather than with any single barrier.
Repair Control SUPPORTED — Apple control over genuine-parts economics. The AppleCare+ motive claim is NOT ESTABLISHED and is not asserted.
Controlled Interfaces STRONGLY SUPPORTED, historical. Lightning / MFi interface control is retained in the evidence record; current burden substantially reduced following USB-C and not converted into a present-tense claim.
Behavioral Manipulation MODERATE — 3.25 / 5. First-party commercial prompting and limited consent/choice-architecture findings.
Information & Privacy MODERATE — 3.25 / 5. Specific first-party advertising and consent concerns despite unusually strong privacy architecture.

Only patterns materially relevant to the final verdict are listed. Patterns investigated and not established — including Artificial Scarcity and Infinite Engagement — appear under Supported vs Not Established rather than here.

Evidence & methodology

Research statusRevised Aug 2026
Evidence coverageComplete
CHI methodologyv2.0
CHI68
CVI84
CFS+16

The Apple investigation included multiple dedicated research passes, explicit falsification testing and a separate adversarial review of the customer-value score.

Historical conduct was preserved in the evidence record but removed from current severity where materially remediated. Evidence labels — Fact, Strongly supported, Analytical inference, Not established, Regulatory — carry the same meanings as elsewhere in the index and preserve the dossier's evidentiary distinctions.

The August 2026 revision was a targeted update, not a rebuild: it investigated the launch of Apple Upgrade and re-tested the existing findings for staleness. Previous findings were left in place where the new evidence did not disturb them.

CHI 68/100, CVI 84/100 and CFS +16 are the settled analytical values for this assessment and are not presented as a range. Overall classification: Customer-favorable — with substantial dependency.

Final verdict

CUSTOMER-FAVORABLE — WITH SUBSTANTIAL DEPENDENCY

CHI68/100Substantial dependency
CVI84/100Exceptional
CFS+16Customer-Favorable

Apple does not need to make its products impossible to leave. It makes them increasingly useful to combine. That creates value. It also creates dependency.

As customers accumulate devices, services, data and relationships, Apple becomes progressively more difficult to replace.

Since August 2026 it also offers to keep them on a lease. Nothing in that lease is priced above sticker — but at the end of the term, ownership is the option the customer has to ask for.

Apple creates extraordinary customer value by making its products work better together.
The more Apple gives you, the harder Apple becomes to replace.