Why an accessible website is better for everyone, and for your ranking

A while back a client got an email from a customer who couldn't finish checking out. Not because the payment failed, but because they navigate the web with a keyboard, not a mouse, and the checkout button never received focus when they tabbed through the page. The order sat abandoned. Nobody had done anything wrong on purpose. Nobody had thought about it at all.

That's what most accessibility problems look like in practice. Not a dramatic failure, just a small gap that quietly costs a business a customer, and quietly tells Google the page is harder to understand than it should be. The two are more connected than most people realise, and so is the cost of getting it right.

What accessible actually means

It's not just alt text on images, though that's part of it. In practice it means: every interactive element reachable with a keyboard, in a sensible order. Text with enough contrast against its background to read in bright light or with tired eyes. Forms where every field has a clear label a screen reader can announce. Buttons that say what they do, not just 'click here'. Video with captions. Headings that follow a logical structure instead of jumping around for visual effect.

None of that is exotic. It's mostly discipline in how a site gets built, not a separate feature bolted on afterwards.

Why it costs almost nothing when it's built in from the start

Semantic HTML, correct heading order, labelled forms, keyboard-friendly navigation: these are decisions a developer makes while writing the code anyway. Choosing a button element instead of a styled div, or writing a proper label instead of a placeholder, doesn't take longer. Checking colour contrast during design takes a few minutes with the right tool, before a single line of code exists. Done at this stage, accessibility is barely a line item. It's just how the work gets done.

Where the money actually gets wasted

The expensive version always happens after the fact.

• Retrofitting a site that was built with visual-only components: a menu that only opens on hover, a slider with no keyboard control, forms built from unlabelled div containers. Fixing this usually means rebuilding the component, not patching it.
• Accessibility overlay widgets: the little plugins that promise instant compliance with one line of code. Most only adjust surface things like font size, and can't fix a page that's structurally broken underneath. Businesses often pay for one of these, then still have to pay to fix the real problem later.
• A redesign triggered by a complaint or an audit, months or years after launch, when the fix now touches navigation, checkout, every form and every page template instead of one component in a design file.
• Legal exposure. Accessibility requirements for businesses selling online in the EU have been tightening, and a formal complaint about an inaccessible site is a far more expensive conversation than a design review would have been.

In every one of these cases, the underlying work isn't more difficult than doing it right the first time. It's just being done twice: once badly, once properly.

Why the same fixes help your ranking

Search engines don't see a page the way a person does. They read structure: headings in order, descriptive link text, alt text that actually describes an image, a logical flow from top to bottom. That's almost exactly the same information a screen reader needs. A page built to be understood by someone who can't see the layout is, almost by definition, a page Google can understand better too. Fast-loading, well-structured, clearly labelled pages tend to rank better for the same reason they're easier to use: there's less ambiguity for anyone, human or machine, trying to make sense of them.

The one question worth asking before you commit

Before signing off on a new website or app, ask whoever is building it: is accessibility something you're designing in from the start, or something we're planning to test for and fix afterwards? The answer tells you almost everything about what this project is going to end up costing, twice.