All insights
Accessibility11 August 20262 min read

Accessibility is a constraint, not a checklist

Treated as an audit at the end, accessibility becomes expensive and grudging. Treated as a design constraint at the start, it mostly just makes the product better.

There are two ways to arrive at an accessible product. One is to design freely, ship, run an audit, and then spend a quarter retrofitting contrast, focus states, labels, and keyboard paths into a system that assumed none of them. The other is to treat accessibility the way you treat performance or security — a constraint that shapes decisions while they're still cheap to make.

The first route is where accessibility gets its reputation as a tax. The second is where it stops being a separate activity at all.

What early constraints actually look like

Deciding accessibility up front is not abstract. It means specific, early choices:

  • Choose a palette that already passes. If a brand color can't carry white text at small sizes, decide that during the palette work — not per-component, six months later, in an argument.
  • Give focus a visible design. Every interactive element needs a focus state that was drawn on purpose, because a keyboard user's cursor is the only cursor they have.
  • Never let color be the only signal. A status that's only red or only green is invisible to some readers and meaningless in a screenshot.
  • Write the label before the icon. If a control can't be named in a few words, the problem is the control, not the label.
  • Set a type floor. Decide the smallest size and lightest weight allowed, and hold it. Most illegible interfaces got that way one 11px caption at a time.

Nearly every accessibility fix is also a usability fix for someone who is simply tired, rushed, or on a bad screen in bright sun.

The overlap nobody talks about

Accessible design is often framed as work done for a small group. In practice the beneficiaries are much broader, because permanent, temporary, and situational limitations produce the same requirements.

Sufficient contrast helps a low-vision reader — and anyone outdoors. Clear focus states help a keyboard-only user — and every power user who never touches a mouse. Captions help a deaf viewer — and the person watching in a meeting with the sound off. Generous touch targets help someone with a tremor — and everyone on a train.

Standards are the floor, not the goal

Contrast ratios and success criteria matter, and we hold to them. But a product can pass every automated check and still be miserable to use: a form that's technically labelled but announces nonsense, a modal that traps focus in the wrong order, a data table that's compliant and unreadable.

So the standard is the floor. The real test is whether someone can complete the task — with a keyboard, with a screen reader, at 200% zoom, on a phone in sunlight. That's not an audit. That's just trying it.

The cheapest test we know

Put your mouse down for ten minutes and use your own product with the keyboard alone. You'll find more real problems in that time than in most reports — and you'll find them while they're still easy to fix.

Working on something this touches? We're usually happy to talk it through before anyone commits to a project.

Start a conversation
Keep reading
Experience Design

Complexity isn't a feature

Enterprise software often wears its difficulty as a badge of seriousness. The complexity belongs in the system — not in the interface people have to operate.

3 min read
Digital Transformation

Digital transformation fails in the handover

Most transformation programs don't fail at strategy or at build. They fail in the gap between launch and the way people actually work on a Tuesday.

2 min read
Digital Products

The dashboard nobody opens

Most dashboards answer questions nobody asked. Designing for decisions rather than for data is the difference between a screen people check and a screen people close.

2 min read

Ready to build what's next?

Let's turn your next business challenge into a digital experience that works.

Let's Talk