Accessibility-First Design
Build accessibility into the foundation — not the finish.
01 — TL;DR
Two sentences.
Accessibility-First Design treats user disability requirements — perceivable, operable, understandable, robust — as foundational constraints from the first design decision, not as compliance items retrofitted after launch. Retrofit accessibility costs 5-10x more than building it in from the start, and the Web Content Accessibility Guidelines (WCAG) provide the structural framework for encoding these requirements into every layer of the product.
This is not a new practice — it is a reorientation of an existing obligation. The legal landscape has made this explicit: the UK Equality Act 2010, the US Americans with Disabilities Act (ADA), and the EU European Accessibility Act all mandate accessible digital products. Accessibility-First Design simply moves compliance from a post-launch audit finding to a pre-design constraint, where it is cheaper, more effective, and produces better products for everyone.
02 — When to Use
Apply this when...
When NOT to apply it
There is no situation where accessibility does not apply. The scope and rigour may vary — a throwaway internal prototype has different requirements than a public-facing product serving millions — but the foundational practice of considering perceivability, operability, understandability, and robustness applies to every design decision. Skip it and you are building exclusion into the product at a structural level.
03 — How It Works
The mechanism
Accessibility-First Design works by reversing the default sequence. Instead of designing, building, auditing, and then fixing, it encodes accessibility requirements as design constraints before the first wireframe is drawn. The cost difference is not marginal — it is structural.
Floor, not ceiling
WCAG compliance is not the same as accessible design. WCAG defines the floor — the minimum requirements that a product must meet. Truly accessible design goes further: it involves disabled users in research, it designs for the full spectrum of assistive technology use, and it treats accessibility as a quality dimension that improves continuously. Passing a WCAG audit means you have met the legal minimum. It does not mean your product is actually usable by the people it claims to serve.
04 — Real Example
GOV.UK Design System — native HTML semantics as baseline
The GOV.UK Design System serves over 60 million users across UK government services. The foundational design decision was to use native HTML elements as the baseline for every component — not custom JavaScript widgets with ARIA bolted on. This single architectural choice meant that keyboard navigation, screen reader compatibility, and focus management worked out of the box, because the browser already handles these behaviours for native elements.
The key decision was to choose native elements over custom ones. Where a custom component was unavoidable, the team documented the full accessibility specification — keyboard interactions, ARIA roles, states, and properties — before writing a single line of CSS. The result is a design system that is accessible by default rather than accessible by remediation, and the maintenance cost of accessibility is close to zero because the browser does the heavy lifting.
05 — In the Wild
Test yourself & see real examples
No examples yet — be the first.
Spotted a product that treats accessibility as first-class — or one where no disabled user was involved in the design process? Submit what the team did right, or what they missed. Every approved example gets attributed to you.
Seen Accessibility-First Design violated in a real product? Help grow the evidence base.
06 — Common Mistakes
Where teams go wrong
07 — Variations & Related Principles
Connected ideas
Accessibility-First Design operates at the intersection of several related but distinct frameworks. Each approaches the problem of inclusive product design from a different angle — philosophy, process, architecture, or cognition.
The most important pairing for Accessibility-First Design is with assistive technology testing. No amount of specification, WCAG knowledge, or automated tooling replaces the insight you get from watching a screen reader user navigate your product — or from navigating it yourself with a keyboard only. The framework provides the constraints; AT testing provides the ground truth.
08 — 10-Min Exercise
Run it right now
⏱ 10 minutes · Solo · No prep
Pick one page or screen in your current product. Run through five checks, spending two minutes on each.
1. Contrast. View the page in greyscale (most operating systems have an accessibility filter for this). Can you distinguish every interactive element, every piece of text, every state change? If anything disappears or becomes ambiguous, you have a colour contrast failure that affects users with low vision and colour vision deficiency.
2. Keyboard navigation. Put your mouse in a drawer. Navigate the entire page using only the Tab key, Shift+Tab, Enter, and Space. Can you reach every interactive element? Can you tell where focus is at all times? Can you escape every modal, dropdown, and menu? If you get stuck or lost, keyboard users and screen reader users cannot use your product.
3. Images and icons. Cover every image and icon on the page with your hand or a sticky note. Does the page still make sense? If any meaning is lost, those images need text alternatives. Decorative images should have empty alt attributes; meaningful images need descriptive alt text.
4. Error states. Submit every form on the page empty, or with invalid data. Are errors described in text (not just colour)? Are they associated with the correct field? Does focus move to the first error? If not, screen reader users will not know what went wrong or how to fix it.
5. Count your failures. Each failure you found in two minutes of looking is a failure that exists for every user who depends on that modality. Multiply by the number of pages in your product. That is your accessibility debt.