Apple's App Store Economics Under Pressure

The most valuable line in Apple’s income statement is one Apple refuses to print.

Apple App Store commission economics are now the regulatory battleground, and the prize is hidden. Apple does not break out App Store revenue. It sits folded inside Services, the category that posted a 76.7% gross margin in the quarter ended March 28, 2026 (Apple Form 10-Q, Q2 FY2026). That margin is the prize. It is also the exact slice that regulators on two continents are now prying open. The same take-a-cut model, applied to rides and deliveries instead of app sales, is dissected in how Uber makes money.

The thesis in one line: the App Store is the highest-margin component of Apple’s Services engine, and the EU’s Digital Markets Act, the Epic injunction, and the pending US DOJ case are rewriting the commission structure that built that margin.

This piece reads the pressure through the filings and the court record, not the keynote framing. Every number ties to a specific SEC filing, regulatory release, or docket. The framing is analytical: how to think about the position, not what to do about the stock.

Key takeaways

• Services is the margin engine, and the App Store is its richest line. Apple Services posted a 76.7% gross margin in Q2 FY2026, an all-time high (Apple Form 10-Q, quarter ended March 28, 2026).
• The EU split a flat toll into a patchwork. As of January 1, 2026, the EU model lands near 7–17% blended depending on distribution path, against the 30% standard rate elsewhere (Apple Developer, DMA resources, 2026).
• The €500M fine set the precedent, not the dent. The €500 million anti-steering penalty is small against $109B of FY2025 Services revenue, but it ratified that anti-steering rules are enforceable (European Commission IP/25/1085; Apple Form 10-K, FY2025).
• A court now sets the US external-transaction rate. The Ninth Circuit affirmed the contempt finding on December 11, 2025, and the Supreme Court declined a stay on May 4, 2026 (Opinion 25-2935; Docket 25A354).
• The structural case is still live. The DOJ suit survived a motion to dismiss on June 30, 2025 and is in contested discovery as of May 2026 (Case No. 2:24-cv-04055).
• The risk and the strength share a quarter. Services printed a record margin in the same quarter the toll was being pried open. Both are true at once.

The margin engine under siege

Services is no longer the sidecar to iPhone. It is the profit story.

Apple’s FY2025 Services revenue was $109,158M, up 14% year over year, at a 75.4% gross margin (Apple Form 10-K, FY2025, year ended September 27, 2025). Then Q2 FY2026 set an all-time high: $30.98 billion in Services revenue at a 76.7% gross margin, up 20 basis points sequentially (Apple Form 10-Q, quarter ended March 28, 2026).

The trend is the point. Services gross margin climbed from 70.8% in FY2023 to 73.9% in FY2024 to 75.4% in FY2025, a 4.6 percentage-point gain over three years (Apple Form 10-K, FY2025). On a category this large, every point of margin is real money.

The App Store is the part of that mix built on the thinnest cost base: a commission on transactions Apple does not fulfill, plus licensing for default placement. That structure is what regulators are targeting. The mechanics of how Services props up the whole device franchise are unpacked in how Apple Services became the margin engine inside iPhone. Here the question is narrower: what happens to that margin when the toll booth is forced open.

How does App Store regulation affect Apple’s commission economics?

App Store regulation moves the commission from a number Apple sets to a number others set. The EU DMA forced alternative distribution and a layered 7–17% fee, the Epic injunction made external-link steering lawful in the US, and a court now decides the external-transaction rate. The posted 30% can stay while the blended rate falls.

That distinction, between the headline rate and the blended rate, is the whole of the economics. A commission is only worth its posted percentage if every transaction has to pass through it. Each regulatory action removes a transaction from the toll, or lowers the toll on the transactions that remain. The aggregate effect is a take rate that drifts down even when the sticker price does not move.

How do alternative app stores work under the EU DMA?

Alternative app stores in the EU must have their apps notarized by Apple, an automated-plus-human review that screens for malware and egregious fraud, replacing the App Store’s full review. Marketplaces also pay Apple a per-first-annual-install fee on their own marketplace app. Multiple stores are live as of February 2026, including Aptoide and AltStore.

The Digital Markets Act did something no prior settlement managed. It forced Apple to allow app distribution outside the App Store entirely.

As of February 2026, multiple alternative app stores operate in the EU, including Aptoide, AltStore, mobinvention’s marketplace, and Onestore (Apple Developer support resources; TechCrunch, February 22, 2026). Apps in those stores skip the App Store’s full review and instead pass through Apple’s notarization: automated checks plus human review to screen for malware, security threats, and egregious fraud.

Apple did not give up the fee. It restructured it. Effective January 1, 2026, the Core Technology Commission (CTC) replaced the earlier Core Technology Fee for all EU business models, set at a 5% baseline plus acquisition and store-services fees (Apple Developer, DMA updates, June 2025 and January 2026 implementation).

The result is a layered toll where there used to be a flat one. EU developers using alternative distribution, or the App Store under the Small Business Program, can land near 10% on subscriptions after year one, against the standard 30% (Apple Developer, DMA apps in EU, January 2026 transition).

The strategic read: Apple traded a clean 30% for a complicated 7–17%. Compliance bought back some of the margin, but not all of it, and not the simplicity.

The €500 million fine and anti-steering compliance

The first hard number the DMA put on the board was a penalty.

On April 22, 2025, the European Commission fined Apple €500 million for violating the DMA’s anti-steering requirements (European Commission IP/25/1085). The finding: Apple prevented developers from telling users about cheaper purchase options outside the App Store, and blocked external links to those options. Apple was ordered to remove the restrictions.

Anti-steering is the quiet load-bearing wall of the commission model. If a developer can route a paying user to a web checkout the moment they are inside the app, the 30% has nothing to attach to. The fine is small against $109B of annual Services revenue. The precedent is not.

Epic Games v. Apple: the injunction and steering freedom

The US version of the steering fight ran longer and ended harder for Apple.

On December 11, 2025, the Ninth Circuit affirmed the district court’s contempt finding, ruling Apple willfully violated the earlier anti-steering injunction (Ninth Circuit Opinion 25-2935). The court upheld permanent injunctions barring Apple from collecting fees on third-party storefront purchases and from blocking external links or payment interfaces.

Apple asked the Supreme Court to stay the order. On May 4, 2026, the request was denied (US Supreme Court, Docket 25A354, Justice Kagan). The case returned to district court to determine what revenue-share percentage, if any, Apple may charge on external transactions.

That last clause is the whole ballgame. The court did not abolish Apple’s right to a cut. It moved the number from “30%, decided by Apple” to “to be determined, by a judge.” For a margin built on a self-set toll, having the toll set externally is the structural change, not the headline.

The US DOJ antitrust case: ongoing litigation

The broadest threat is also the slowest.

The DOJ’s antitrust suit against Apple, filed March 21, 2024 in the District of New Jersey (Case No. 2:24-cv-04055), attacks the iPhone–App Store ecosystem lock itself, not just the steering rules. On June 30, 2025, the court denied Apple’s motion to dismiss, sending the case into discovery. As of May 2026, Apple has been disputing its discovery obligations with the DOJ, a sign of active, contested litigation rather than a settlement glide path.

A motion-to-dismiss denial is not a verdict. It does mean a federal judge found the government’s theory plausible enough to test on the evidence. That alone keeps the structural question open: whether the integration that makes the App Store a single mandatory toll booth is itself the violation.

The Commission Pressure Ledger: scoring each regime

The clearest way to see Apple App Store commission economics under regulation is to put the regimes in one frame and score what each one actually takes off the toll. Call this the Commission Pressure Ledger: a framework that rates a platform’s commission position not by its posted rate but by four properties that decide the blended rate. The posted commission, whether steering out is allowed, whether alternative distribution exists, and the direction of pressure on the blended take. The ledger is an original synthesis of Apple’s published fee documentation across jurisdictions; the underlying rates are cited in the sources block. The point of naming it is reuse: you can run any take-rate platform, Apple’s or a marketplace startup’s, through the same four columns.

Regime	Posted rate	Small Business / reduced	Steering allowed?	Alternative distribution?	Blended-rate pressure
Worldwide (pre-DMA baseline)	30%	15% (under $1M/yr)	No	No	None (toll fully intact)
US, post-Epic injunction	30% in-app	15% (Small Business)	Yes, external links	No (App Store only)	Down, via external checkout leakage
EU, post-DMA (Jan 2026)	~17% blended (2% acquisition + 5–13% store services + 5% CTC)	~10% (Small Business)	Yes	Yes (Aptoide, AltStore, others)	Down hard, via rate cut + distribution loss

Sources: Apple Developer App Store fee documentation, 2026; Apple Developer DMA resources, January 2026; Ninth Circuit Opinion 25-2935.

Read the rightmost columns. Outside the EU, Apple still owns distribution and now only loses the steering wall, so the pressure is a slow leak. Inside the EU, it loses distribution exclusivity and the flat rate at once, so the pressure is structural. The commission did not collapse. Run it through the ledger and the mechanism is plain: each “Yes” in the middle columns subtracts transactions from the toll, and each one bends the blended rate down regardless of what the posted number says.

The commission structure redrawn

Put the three regimes side by side and the fragmentation is the story.

Outside the EU, Apple still owns distribution and now only loses the steering wall. Inside the EU, it loses distribution exclusivity and the flat rate at once. The commission did not collapse. It split into a jurisdiction-by-jurisdiction patchwork, and patchworks are where blended margins quietly leak. That same self-set-toll-meets-distribution dynamic runs in reverse through Google’s AI strategy as a distribution war, where owning the surface is the defense rather than the liability.

Methodology: how to read the margin exposure

A note before the takeaway, because the temptation is to model a precise hit. Apple’s disclosures do not support that, so this stays directional.

• Inputs: Services gross margin of 76.7% (Form 10-Q, Q2 FY2026) and Services revenue of $30.98B that quarter; the EU’s blended commission range of roughly 7–17% versus a 30% standard rate.
• Assumption: the App Store is the highest-margin line inside Services, because its incremental cost (hosting a third-party transaction, licensing default placement) is near zero. Apple confirms it does not separately disclose App Store revenue (Form 10-K, FY2025), so this is a qualitative ranking, not a quantified one.
• Sensitivity: the margin effect scales with how much App Store volume sits in DMA-affected jurisdictions and with how far the post-Epic external-transaction rate lands below 30%. Both are unknown from public filings.
• What this misses: Apple does not break out App Store revenue, EU revenue, or commission revenue. Any single-number estimate of the margin hit would be invented. The honest output is a direction (down) and a mechanism (forced rate compression plus distribution leakage), not a figure.

Where the margin is vulnerable

A credible read names the soft spots. There are three.

The blended rate, not the headline rate, is what compresses. The EU model lets the most price-sensitive, highest-volume developers reach the lowest tiers. If the apps that generate the most transaction value migrate to 10% or to alternative stores, the average commission falls even while the posted 30% stays on the page for everyone else.

Steering freedom unbundles the payment from the platform. Once a developer can link out to a web checkout, Apple’s cut applies only to the transactions that stay in-app. The post-Epic US outcome makes external links lawful nationwide. The remaining lever is the to-be-determined external-transaction fee, and a court sets it, not Apple.

Compliance is now a recurring cost line, not a one-time fix. Notarization infrastructure, jurisdiction-specific fee engines, a €500 million penalty, and an active DOJ discovery fight are ongoing operating reality. They do not show in a clean line item, which is precisely why they are easy to underweight. Why a single point of gross margin compounds into a very different business is the throughline of why gross margin is destiny in SaaS.

None of this is fatal on today’s evidence. Services still printed a record 76.7% margin in Q2 FY2026 (Form 10-Q). The pressure is real and the margin is intact, in the same quarter. That tension is the entire story.

The bear case: what the skeptics get right

The strongest argument against the margin-under-pressure thesis is the inverse one: that the pressure is real but immaterial, and the worry is mostly noise.

The bear case runs like this. The EU is a single region. The €500 million fine is a rounding error against $109B of FY2025 Services revenue (Apple Form 10-K, FY2025). The post-Epic external-link freedom assumes developers will build, maintain, and market their own web checkout and that users will abandon the friction-free in-app purchase to save a few percent. Most will not. Apple still controls notarization, the payment sheet, the default placement, and the discovery surface, so even an “alternative” marketplace lives downstream of Apple’s plumbing. On this reading the toll booth was never the moat; the integration and the trust were, and neither is going anywhere. The record margin in Q2 FY2026 is the bear case’s strongest exhibit: after a fine, an injunction, and a live DOJ case, the margin went up.

There is precedent the skeptics can point to. Anti-steering and alternative-payment options have existed in some form for years in markets like the Netherlands and South Korea, and the practical migration of paying users away from in-app purchase has been modest. Behavioral inertia at the checkout is real and verifiable, and it favors the incumbent that owns the default.

Here is the honest weighing. The bear case is right that the near-term damage is small and that integration, not the commission alone, is the durable asset. It is weaker on direction and on who holds the pen. A self-set toll that becomes a court-set or regulator-set toll is a different kind of asset, even if the first year’s revenue barely moves, because the optionality to raise or hold the rate is gone. The blended rate only has to leak slowly to matter on a line this large and this high-margin. The bear case is a reason to discount the headlines, not a reason to ignore the structural change underneath them. Watch the Services gross-margin trend quarter over quarter; that is where slow leakage would surface first.

What operators should take from this

If you run a platform that takes a cut, the App Store is the case study in what happens when the cut becomes contestable. The transferable lesson is that a self-set toll is a policy choice protected by control over distribution, not a law of nature. Apple held 30% for over a decade because it owned the only on-ramp. The moment regulators forced a second on-ramp and banned blocking the exits, the rate became negotiable, and negotiable rates trend toward marginal cost.

Here is the playbook, five concrete moves you can run against your own take-rate business this quarter:

1. Score your own platform on the Commission Pressure Ledger. Take the columns above (posted rate, reduced tier, steering allowed, alternative distribution, blended-rate pressure) and rate your platform honestly. If you depend on being the only on-ramp and the only checkout, you are the pre-DMA row, and you are one regulator or one large unhappy partner away from the EU row.
2. Price the cut against the switching cost, not the value delivered. Apple’s 30% was anchored to lock-in. When the lock-in eroded, the anchor went with it. If your take rate depends on being the only path, model what it survives when a second path opens, before someone opens it for you.
3. Own the relationship, not just the rail. The developers fleeing to alternative stores and web checkout were never loyal to the App Store; they were captive to it. Captivity is not retention. Convert the captive into the chosen while you still set the terms, the same surface-ownership logic that powers Microsoft Copilot and enterprise lock-in.
4. Treat compliance as a permanent cost center, and price it in early. A patchwork of jurisdiction-specific fee logic, notarization, and discovery defense is now part of Apple’s cost base. Any platform operating across regulatory regimes should carry that as a standing line, not a surprise.
5. Watch the blended take rate, not the posted one. For Apple, the number to track is the Services gross-margin trend, because that is where rate compression and distribution leakage surface first. For your own platform, instrument effective take rate (fees collected divided by gross transaction value) and put it on a dashboard, because the posted rate will lie to you long after the real one starts falling.

To see how a healthy take-rate engine reads when it is not under siege, the structural pricing tradeoff is laid out in usage-based pricing versus seat-based pricing.

Apple’s App Store will remain enormously profitable. The question the filings and dockets pose is narrower and sharper: not whether the toll booth survives, but who gets to set the toll. For more than a decade the answer was Apple. As of 2026, the answer is increasingly a court.

---

Analysis, not investment advice. Figures are drawn from Apple Inc.’s public SEC filings (Form 10-K FY2025 and Form 10-Q Q2 FY2026), European Commission and US court records, and cited inline by source and period. Frameworks here are for understanding business strategy and tradeoffs, not for making buy or sell decisions.

Want the full toolkit for reading filings like this, the segment-margin worksheet, the take-rate-under-pressure framework, and the Commission Pressure Ledger used above? It’s in the Tech Business Analysis Playbook.