A recent discussion ignited by prominent web developer and W3C contributor Lea Verou has brought to the forefront a critical, albeit often overlooked, aspect of modern web design: the implementation of light and dark mode toggles. Verou’s critique, shared across her blog and social media platforms, challenges the common practice of offering three distinct state options — Light, Dark, and System — arguing that a well-designed two-state control can not only express all necessary user preferences but also significantly enhance user experience by reducing cognitive load and UI clutter. This perspective is gaining traction among developers and UX professionals who are increasingly scrutinizing default implementations for greater efficiency and user-centricity.
The Evolution of Dark Mode and System Preferences
The concept of dark mode, offering a user interface with a dark background and light text, has a history stretching back to early computing interfaces, often driven by the limitations of cathode-ray tube (CRT) monitors and command-line interfaces. However, its widespread adoption in mainstream consumer software and operating systems gained significant momentum in the late 2010s. This resurgence was fueled by several factors: perceived benefits for eye strain, especially in low-light environments; potential for battery saving on OLED screens; and aesthetic preferences.
As dark mode gained popularity, operating systems like Apple’s macOS (Mojave, 2018), Google’s Android (Q, 2019), and Microsoft’s Windows (10, 2018) began integrating system-wide dark mode settings. Concurrently, web standards evolved to allow websites to respect these system-level preferences. The prefers-color-scheme media query in CSS, standardized by the W3C, became the technical cornerstone, enabling developers to style their websites dynamically based on the user’s OS theme setting. This innovation was a significant step towards a more harmonious and integrated user experience across the digital ecosystem.
However, with this newfound capability came a design dilemma. Many websites, aiming to provide users with maximum control, began implementing theme toggles that offered not just "Light" and "Dark" options, but also a "System" option. This "System" choice explicitly tells the website to defer to the operating system’s setting, mirroring the browser’s default behavior if no site-specific preference is set. It is precisely this "System" option in persistent, site-specific toggles that Lea Verou contends is largely redundant and, in many cases, detrimental to good user experience.
Deconstructing the Tri-State Toggle: A Question of Cognitive Load
Verou’s central argument, articulated through a succinct visual illustration and detailed blog post, posits that displaying a "System" option unnecessarily complicates the user interface. She emphasizes that the user’s default expectation is for a website to respect their operating system’s color scheme preference. If a user has no explicit site-specific preference, the browser, by default, will apply the styling dictated by prefers-color-scheme. This means the "System" state is effectively the initial default state.
When a user interacts with a theme toggle, their primary intent is typically to override the current display if they dislike it. If their system is set to Light mode, and they encounter a website that is also Light by default, they might toggle it to Dark. This action, Verou argues, should be sufficient to store their preference (e.g., in localStorage) for future visits. If they later toggle it back to Light, that action should also be stored as an explicit preference. The core idea is that the "System" state is the baseline, and any user interaction implies an override, whether to Light or Dark.
Consider the user’s journey:
- First Visit, No Override: The user visits a website. Their OS is in Dark mode. The website, respecting
prefers-color-scheme, displays in Dark mode. No user interaction is required. - First Visit, Override: The user visits the website. Their OS is in Dark mode, but they prefer Light for this specific site. They click a "Light" option or toggle. This preference is stored.
- Subsequent Visit, Override Respected: The user returns. The stored preference for Light mode is applied, irrespective of their OS setting.
- Toggling Back: If the user later decides they want Dark mode for the site, they toggle it. This new preference is stored.
In this flow, a separate "System" button doesn’t add functional value for a persistent toggle. The system preference is the implicit default until an explicit override is made. Presenting "Light," "Dark," and "System" forces users to make a decision about a state that is already implicitly handled. This introduces cognitive load, however minor, which accumulates across numerous interactions. As Chris Coleman, a user on Bluesky, succinctly put it in the thread surrounding Verou’s post: "All I ever wanted was for my OS to be dark, not every website I look at. It was a huge leap by the browser vendors to tie all web content to that system preference." This highlights a user expectation for browser-level control and site-level adherence unless explicitly overridden.
The Power of a Good Two-State Control
Verou’s proposed solution involves a simple two-state toggle, often represented by an icon (e.g., a sun/moon) or a switch that alternates between "Light" and "Dark." The key lies in its underlying logic:
- Initial Load: If no user preference is stored locally (e.g., in
localStorage), the website defaults to the system’sprefers-color-scheme. - User Interaction: When the user clicks the toggle, it switches to the alternative mode (e.g., from Light to Dark, or Dark to Light). This explicit choice is then stored in
localStorageas their preferred theme for that specific site. - Persistence: On subsequent visits, the stored preference overrides the system setting.
This approach aligns with well-established UX principles that advocate for minimalism and intelligent defaults. By removing the "System" option, the interface becomes cleaner, and the user’s interaction path is simplified. The control is tangential to the user’s primary goal on the website; therefore, minimizing its complexity contributes to a smoother overall experience. The perceived "loss" of the "System" option is negligible because the system preference is the baseline behavior when no explicit site-level choice has been made. If a user wants to revert to following their system preference, they simply choose the "Light" or "Dark" option that matches their current system state, effectively pinning that state, even if the OS later changes. The user can always change their mind with a single click if their OS setting diverges from their site-specific override.
Expert Perspectives and Industry Reactions

The discussion sparked by Lea Verou has resonated deeply within the web development and UX communities. Many practitioners echo her sentiment, recognizing the subtle but significant impact of UI choices.
A hypothetical statement from a lead UX designer at a major tech company might articulate: "Our goal is always to minimize friction and maximize intuition. Lea’s analysis on dark mode toggles is spot-on. When a system provides a default, the user expects that default to be respected unless they actively choose otherwise. Adding a ‘System’ button on every site-level toggle adds an unnecessary layer of decision-making that detracts from the core content or functionality the user came for. We are constantly striving for ‘zero-UI’ where possible, letting the user’s environment guide the experience."
From a web developer’s standpoint, simplifying the logic also offers benefits. A senior front-end developer might observe: "Implementing a tri-state toggle, while seemingly offering more control, often introduces more complex state management and potential edge cases. A two-state model, backed by prefers-color-scheme and localStorage for overrides, is generally simpler to code, test, and maintain. This allows us to focus more on core functionality and performance, rather than intricate UI decisions that don’t add real value for the user."
The conversation also touches on the broader vision for how user preferences should be handled across the web. The idea, supported by figures like Chris Coleman, is that the ultimate control over color schemes should reside at the browser level, allowing users to dictate their preference once and have it consistently applied everywhere, with site-specific overrides as exceptions, not the rule. This aligns with the long-term goals of web standards bodies like the W3C, which continually work towards empowering users through browser-level settings and simplifying developer responsibilities.
When a Tri-State Control Is Appropriate
While Verou strongly advocates against the persistent tri-state toggle in typical website headers, she acknowledges that there are specific scenarios where a three-state control (or more) is indeed appropriate and beneficial:
-
Global Preference Settings: When the control manages a theme preference that applies across multiple applications, services, or an entire operating system. For example, the dark mode setting within an operating system’s display preferences, or a global theme setting within a large platform like Microsoft 365 or Google Workspace, which affects numerous integrated applications. In these contexts, the "System" option allows users to align their preference with a broader, overarching system setting, distinct from individual application overrides.
-
Theme Builder or Customization Interfaces: In applications or websites where the primary user intent is to customize the appearance, such as a website builder, a design tool, or a personal profile customization page. Here, the user is actively engaged in selecting and refining visual themes, and a wider range of options, including "System," "Light," "Dark," and potentially custom themes, provides necessary granular control. The user is in a "creation" or "configuration" mode, where explicit choice is expected and desired.
These distinctions are crucial. The problem arises when a tri-state toggle, designed for broad system-level control or dedicated customization, is indiscriminately applied to every persistent UI element on a standard content-driven website.
Broader Implications for Web Design and Development
Lea Verou’s detailed critique and the subsequent discussion carry significant implications for the future of web design and development:
- Enhanced User Experience: By simplifying UI controls, websites become less cluttered and more intuitive. This reduction in cognitive load contributes to a smoother, more enjoyable user journey, fostering greater engagement and satisfaction. It respects the user’s primary intent, which is rarely to endlessly configure display settings.
- Improved Developer Practices: The guidance encourages developers to adopt more thoughtful and user-centric UI design patterns. It promotes adherence to web standards (like
prefers-color-scheme) as the default behavior and intelligent use of client-side storage (localStorage) for managing explicit overrides. This can lead to cleaner, more maintainable codebases. - Accessibility Considerations: While not explicitly detailed in the original excerpt, consistent and predictable theme management is crucial for accessibility. Users with visual impairments, light sensitivity, or cognitive disabilities benefit from interfaces that behave predictably and offer clear, unambiguous controls. Reducing unnecessary options helps minimize confusion and potential frustration.
- Future of Web Preferences: The debate subtly pushes towards a future where more user preferences are managed at the browser or operating system level, reducing the need for redundant site-specific controls. This would empower users with greater control over their overall digital environment and free developers to focus on unique application-specific features. The "no toggle at all" scenario, where the site simply respects the system preference without providing any override, might be the ideal for many content-focused websites.
Conclusion
Lea Verou’s incisive analysis of light/dark mode toggles serves as a potent reminder that even seemingly minor UI decisions can have a substantial impact on user experience. By advocating for a streamlined two-state control over the common three-state implementation, she champions principles of minimalism, intelligent defaults, and reduced cognitive load. This approach not only cleans up the user interface but also aligns more closely with user expectations regarding system-level preferences.
As the web continues to evolve, the focus on user-centric design remains paramount. Developers and designers are encouraged to critically evaluate their existing dark mode implementations, considering whether the "System" option truly adds value in persistent toggles or merely introduces unnecessary complexity. Embracing a more refined approach, where system preferences are respected by default and explicit overrides are handled elegantly with a simple two-state control, represents a significant step towards creating a more intuitive, efficient, and user-friendly web. The ultimate goal is an experience where users can seamlessly interact with digital content, unburdened by redundant choices, and where their preferences are implicitly understood and gracefully accommodated.
