The Friction Was Never Symmetric
Meta description: A good UX instinct — don't interrupt the flow of the experience — produces a very different outcome when the action being smoothed is a real-money purchase instead of a game action. A design breakdown of the $520M Epic Games settlement.
In December 2022, Epic Games agreed to pay $520 million to resolve two Federal Trade Commission actions: a $275 million penalty for violating COPPA, the federal law governing children's online privacy, and a separate $245 million consumer-refund fund tied to Fortnite's in-game purchase design — at the time, the largest FTC refund amount in a gaming case and the agency's largest administrative order ever. (Epic settled without admitting or denying the FTC's allegations, standard practice in FTC consent orders; the design details below are drawn from the complaint and the FTC's public description of the case.)
The purchase-design allegations are the more broadly useful half of this case, because the mechanism isn't specific to gaming. It's a UX principle — minimize interruption, preserve flow — applied uniformly to every button in an interface, including the ones that move real money, without a separate design rule for that category of action.
The environment: flow is the product
Fortnite, like most live-service games, is built around momentum. Interrupting a player mid-match with confirmation dialogs, modal pop-ups, or multi-step flows is bad design in almost every other context in the product — it breaks immersion, adds friction to the core loop, and is the kind of thing a UX team is right to minimize. That instinct isn't wrong. It's one of the more defensible, well-established principles in game design, and it's the same instinct that makes checkout flows shorter, onboarding flows leaner, and most consumer software better.
The FTC's complaint describes a button configuration where that same low-friction, minimize-interruption design was applied to purchase actions as well as gameplay actions: charges could reportedly be triggered by pressing a button to wake the game from sleep mode, by an input during a loading screen, or by pressing a button adjacent to the one a player meant to press while simply previewing an item. None of these are described as a deliberately deceptive checkout screen. They're described as ordinary interface buttons that happened to be wired to a real-money transaction, sitting in the same visual and spatial context as buttons wired to free, reversible game actions.
What the design actually did
Two separate mechanisms are worth pulling apart, because they sit on opposite sides of the same transaction and reinforce each other.
No confirmation boundary around spend. In most well-designed purchase flows, there's a deliberate friction boundary — a step that exists specifically to separate "browsing" from "spending," even if it's just one tap. The complaint's account of Fortnite's button configuration describes a purchase path with effectively no such boundary: the same input vocabulary used for gameplay (single presses, quick taps, buttons near other frequently-used buttons) doubled as the input vocabulary for spending, with no distinct visual or interaction context marking the transition from one to the other.
Default-saved payment method removed the remaining friction. The complaint also describes Epic's system as saving the account's associated credit card by default, so that once a card was on file, a single button press was sufficient to complete a purchase — no additional authentication or cardholder confirmation step. For an adult account, that's a convenience trade-off some users would accept. For an account used by a child, which the COPPA half of the case specifically addresses, it removes the one obvious point in the transaction where a parent would ordinarily need to be involved — a physical card, a password, a second device.
The refund path sat behind an asymmetric amount of friction relative to the purchase path. The FTC describes the refund process as requiring users to locate and navigate a comparatively difficult, multi-step path under a "Settings" tab. Whatever the individual reasoning for that placement, the practical effect is the same shape seen in the subscription-cancellation cases elsewhere in this series: the path toward a transaction is short and the path away from an unwanted one is long, even when neither path was necessarily designed with that comparison in mind.
The behavioral and design principle at work
The generalizable finding here isn't "Fortnite has bad UX." It's closer to the opposite: Fortnite's UX was optimized correctly for its stated goal — minimizing interruption — and the goal itself needed an exception for one category of action that the design system didn't carve out. Two specific, well-documented UX/behavioral concepts explain why that gap is so easy to miss:
- Fitts's Law and proximity-driven error, which predicts that closely-spaced, similarly-styled interactive targets increase the rate of unintended selection — exactly the mechanism behind a purchase being triggered by a button meant to preview an item or wake the screen.
- Default effects, one of the most replicated findings in behavioral economics: whatever is pre-selected or requires no action is disproportionately likely to be what happens, regardless of whether it reflects deliberate intent. A saved card isn't a dark pattern by itself — most e-commerce relies on saved cards for good reason — but a saved card with zero additional confirmation step, applied to an audience that includes children, changes what "intent" means at the point of purchase.
Neither of these principles is exotic. They're standard tools in any UX or growth practitioner's vocabulary, applied here — as in the cancellation-flow cases elsewhere in this series — without a rule distinguishing "smooth this interaction because friction only costs us engagement" from "add a deliberate boundary here because friction is the only thing standing between a browsing action and a real-money one."
The trade-off, stated plainly
This is a genuine design tension, not a case where the "right" choice was obvious in advance: minimizing interruption is good for engagement and good for most of the product experience, and a confirmation step on every purchase-adjacent action would measurably hurt conversion and immersion if applied indiscriminately. The mistake available to any team building a game, app, or one-click purchase flow isn't "wanting low friction." It's not drawing an explicit line around which actions get the low-friction treatment and which get a deliberate, separate confirmation boundary — and, separately, not treating "the account may belong to a minor" as a condition that changes where that line sits.
What this means for product and growth teams building purchase flows
- Give real-money actions a distinct interaction grammar. Purchase-triggering buttons should be visually and spatially distinguishable from adjacent free actions — different color language, deliberate spacing, or a brief, non-blocking confirmation state — specifically so the low-friction design principle used everywhere else in the product doesn't apply to this one category by default.
- Treat saved-payment-method defaults as a settings decision, not a technical default. Whether a stored card requires re-authentication above a spend threshold, after a period of inactivity, or for an account flagged as a minor's, is a product policy choice with real behavioral consequences — it shouldn't be inherited silently from whatever the payment SDK does out of the box.
- Design the refund or dispute path to the same standard as the purchase path. If a purchase can be completed in one motion, an unintended purchase should be reversible in a comparably small number of steps — treating refund-path length as its own reviewed UX metric, not an afterthought placed wherever it fits in a settings menu.
None of these require slowing down the core game loop or the parts of the product where low friction is genuinely the right call. They require one explicit exception, deliberately designed, for the narrow set of interactions where friction is doing a job beyond convenience.
This is part of a series on the design and decision-making behind marketing-compliance enforcement actions. See the full series or get in touch if you're building a purchase or upgrade flow and want a second read on where the friction is — and isn't — landing.