Apple's New EU App Store Terms Take Effect October 1. The Commission Cut Is Real, and So Is the 12-Month Lock-In.
Apple's new business terms for EU apps go live on October 1, and developers can sign them today. The headline is a commission cut that is real: link out of your app to take payment and Apple takes 15%, or 10% if you are in the Small Business Program. The Core Technology Fee, the per-install charge that terrified anyone with a free app and a viral week, is gone.
The part I have not seen quoted anywhere is one sentence in Apple's own announcement. You select your payment options and you "must maintain those options for 12 months." That turns a pricing decision into a one-year commitment, and it is the only part of this that should change how you actually decide.
The rates, without the spin
Straight from Apple's newsroom post:
- App Store apps using Apple In-App Purchase: 26%. Drops to 15% for the Small Business Program, the Mini Apps Partner Program, the Video Partner Program, and for auto-renewing subscriptions after their first year.
- App Store apps using alternative payment processing inside the app: 20%, or 10% in those programs.
- App Store apps that link out to complete the purchase: 15%, or 10% in those programs.
- Apps distributed through alternative marketplaces or the web: a 5% Core Technology Commission.
The Core Technology Fee is replaced by that 5% Core Technology Commission. The initial acquisition fee and the store services fee are both eliminated. Apple also now lets you offer Apple In-App Purchase alongside alternative payment options in the EU, which was previously not permitted.
If you are a Small Business Program developer, the practical range you are choosing between is 15% (Apple IAP) versus 10% (link out or alternative processing). That is a five point spread, and five points is not nothing, but it is a lot narrower than the 30% versus 0% framing that people have been arguing about since 2020.
Five points is not the number that matters
Here is where I think most of the coverage goes wrong. It treats this as an arithmetic problem: find the lowest percentage, pick it, bank the difference.
Payment processing is not free. If you link out to Stripe you are paying roughly 2.9% plus 30 cents on card transactions, plus EU VAT handling that Apple was doing for you, plus chargebacks, plus refund handling, plus the support load when someone's card fails at 2am and they email you instead of Apple. Run that against a 5 point gross saving on a €9.99 subscription and you are fighting over something like 30 cents a month per subscriber, before your own time.
At 100 subscribers that is €30 a month to take on a payments stack. At 10,000 subscribers it is €3,000 a month and the calculus flips hard. The threshold where linking out starts to pay is a real number and it is specific to your price point, your refund rate, and how much you value not being a payments company.
The 12-month clause is the actual decision
This is what I would spend the time on. Apple's framing is that the commitment exists "to provide consistency and clarity for users." Fine. The effect on you is that you are picking a configuration in the next week or so and living with it through most of 2027.
That matters because the thing you are trying to predict is your own growth. If you are at 300 subscribers today and you link out because the percentage is lower, you have spent a year building and supporting a checkout flow to save a few hundred euros. If you are at 8,000 and you stay on Apple IAP because it is easier, you have committed to leaving real money on the table for four quarters.
The asymmetry runs one way. Getting it wrong on the side of "stayed with Apple IAP" costs you margin you can calculate. Getting it wrong on the side of "built a checkout I did not need" costs you engineering weeks, a support surface, and the opportunity cost of whatever you did not ship instead. For most solo operators, the second mistake is the expensive one.
What I would actually do
Pull your last twelve months of EU revenue and answer three questions.
First, what is your actual EU share? A lot of people are about to spend a week on this for revenue that turns out to be 6% of their total. If the EU is a rounding error for you, sign the terms, stay on Apple IAP, move on.
Second, at your current price point and volume, what does 5 points actually produce in euros per month? Write the number down. If it is smaller than what you would charge a client for a day of work, the decision is made.
Third, where will you plausibly be in twelve months? Not your best case. Your median case. If the 5 points only pays off in a scenario you would not bet on, do not configure for that scenario.
If the numbers say link out, do it now rather than in February, because the 12-month clock starts when you select. If they say stay, stay, and revisit in October 2027 with a year of better data.
One more thing worth checking if you have an app for kids or teenagers: apps in the Kids category cannot link out to websites for transactions at all, users under 18 require a parental gate for alternative payments or link-outs, and users under 13 cannot be linked out to websites for transactions. If that describes your audience, the choice may already be made for you.
Where I could be wrong
I am treating the 12-month commitment as a serious constraint, and it is possible Apple handles this loosely in practice, or that the European Commission pushes back on the clause before it bites anyone. The DMA process has moved before and could move again. If the lock-in gets softened, the calculus becomes ordinary arithmetic and my caution here looks overcautious.
I am also arguing from the position of a small developer, where the cost of building payments infrastructure is enormous relative to the saving. If you already run a web checkout for a companion product, most of my objection disappears, and the lower rate is close to free money.
But I would rather make that decision deliberately in September than discover in March that I am eleven months into a configuration I picked in an afternoon.
Author
Lukas
@lukcombinator