Material Design: why big companies borrow Google's system

Material Design is an open design system that Google published in 2014. It gives teams shared guidelines, components, and code for building interfaces. Large companies adopt it because a public system saves them work they would otherwise redo on their own.

What Material Design is and the problem it solves

A design system is a written agreement about how an interface behaves. Material Design covers color, typography, shape, motion, and components such as buttons, cards, and navigation bars. The public site, Material Design 3, describes the system as “Google’s open-source design system”.

It solves two problems: duplicated work and teams drifting apart. Without a shared reference, every team writes its own button, its own spacing scale, and its own rule for a disabled state. In the end, people have to relearn the same interface in every product they open.

Why Google built it

Google announced Material Design on June 25, 2014, at Google I/O, with Matías Duarte, then VP of design, on the keynote stage. Before Material, Google ran a web-only refresh called Kennedy. Google’s own history of the Material eras states the goal: make Gmail, Chrome, Android, and the rest feel like one family. That same account records the team starting with six designers and about 200 products, both figures as Google’s own count.

The metaphor was paper. Designers cut icons from real paper to study shadow, and printed layouts on large sheets for review. Motion rules came from watching paper slide and settle. The project had a working name, Quantum Paper.

Google tells this part as the story of opening its work up to everyone. In the same Google account of the history, a designer from the original team recalls that within a year, more than a million products in the Android store used parts of the system.

What changed from Material 1 to 2 and 3

Material 1 ran from 2014 to 2017, built on paper physics, rigid grids, and minimal surfaces. Its mantra was that motion carries meaning.

Material 2 ran from 2018 to 2020 and filled in the gaps. Google’s own history of the eras records more than 20 components drawn and only one or two implemented in code, its own count. Material 2 shipped coded components, tested accessibility around keyboard focus and color contrast, and opened theming. The team’s analogy was a box of Lego: one base, and each brand on top.

Material 3, or Material You, arrived in 2021 with Android 12 and moved toward expression: variable type, eclectic shapes, bold color, and lively motion. The same account says it reached billions of Android phones and tablets, a claim Google makes about itself. Material 3 Expressive followed in May 2025, and Android’s documentation ties it to the Android 16 visual style.

The principles that survive every version

Color, typography, and shape stay the three knobs. Android’s Material 3 documentation says a theme contains exactly those subsystems, and that a change to any of them flows into every component.
Motion explains state. Material 1 drew motion rules from paper physics. Google’s research on Material 3 Expressive lists motion among the fundamental parts of expressive design, with color, shape, size, and containment.
One code base, many brands. Material 2 opened theming so external teams could keep the same base and apply their own identity.

Consistency itself was redefined. A Google Design film series on Material You asks whether consistency has to mean uniformity.

Why large companies adopt a public system

The floor is already built. An adopter starts from components tested across many products, languages, and screen sizes.
Maintenance is shared. Someone else tracks platform changes, accessibility guidance, and new form factors. The adopter pays for the override layer, not the base.
The vocabulary is public. Designers, developers, and new hires point to the same documented names.
The code path is official. Jetpack Compose implements Material 3 for Android, and Flutter made Material 3 the default design language of its widgets in version 3.16.

What you gain and what you lose

You gain speed to a first release, inherited accessibility work, and an interface your users already recognize.

You lose distinctiveness when the theming layer stays thin. Google’s own researchers opened the Material 3 Expressive work in 2022 with a blunt question about why Google apps looked “so similar” and “so boring”. A public system makes that outcome the default.

You also lose control of the roadmap. The move from Material 2 to Material 3 changed the theming model, and teams with deep customizations had to migrate.

When adopting is the wrong decision

When the interface is the product. If the interaction is the differentiator, a shared component library flattens the part that sells.
When platform conventions are the requirement. iOS and Android users hold different expectations, and a system built for one can fight the other.
When nobody can own the override layer. Theming is a second system on top of the first, and an unowned theme rots.
When the fit is wrong from the start. A dense desktop tool and a consumer mobile app have different needs.

The decision fits in one short rule. Adopt when the gain is speed on work that isn’t your differentiator. Build when the interface itself is what customers pay for.

A smaller version of the same decision, measured in returns, is in the voltage post.

Where the research disagrees with the post

Three articles from Nielsen Norman Group cut against claims made above.

NN/g defines a design system as a set of standards that manage design at scale (Design Systems 101). That definition describes the system a company builds and maintains, while this post describes the system a company adopts ready-made, and the cost migrates to the adjustment layer.

The second article states that defending consistency and standards doesn’t mean everything should look and behave the same (Consistency and Standards). The claim above, that a public system makes a uniform and dull result the default, is exactly what NN/g denies.

The third article reports that an attractive product is perceived as more usable, before any task is measured (The Aesthetic-Usability Effect). This post treats visual identity as a cost of singularity, and NN/g gives the visual layer an effect that appears before any measurement.

The guide has enough in it for more than one post. Its axes are broken out into their own pieces: tokens and theming, color and the color role, and typography. Each one goes deeper into a part without repeating this post.