Overlays Won't Save You: The Accessibility Order of Operations for UK SMEs
Photo: Verdy p, CC0, via Wikimedia Commons
Since the European Accessibility Act came into force in June 2025, a lot of UK business owners have received a remarkably similar phone call. The script usually involves a stern mention of legal exposure, a monthly fee, and a snippet of JavaScript that will, allegedly, make the whole problem disappear.
It will not. Not because the vendors are all villains, but because of what an overlay physically is: a script that sits on top of your pages and tries to compensate for the HTML underneath. It cannot rewrite your markup, your contrast ratios or your form errors. It can only paper over them, sometimes badly enough to make things worse.
If the accessibility conversation has landed on your desk and you are wondering where to start, here is the order we would spend money in. Sequence matters here, because the cheap fixes are the ones that remove the most barriers.
First: contrast and text size
Before anything clever, check that people can actually read the page. Body text wants a contrast ratio of at least 4.5 to 1 against its background; large text can get away with 3 to 1. Grey-on-white at 14 pixels is the single most common failure on UK service websites, and it is usually a design decision made in a meeting about how 'clean' something looked.
Bump body copy to 16 pixels minimum, ideally 17 or 18. Give line height some room. Stop using light weights for paragraphs. This costs nothing, takes an afternoon with a contrast checker, and helps everybody — tired eyes, bright sunlight, small laptops and the significant chunk of your audience over 50.
Second: keyboard and focus
Unplug your mouse and try to use your own site. Can you reach the navigation? Can you see where you are as you tab? Is the focus outline visible, or did somebody remove it because it was untidy?
Every interactive control needs a clear focus state, and the tab order should follow the visual order of the page. Modal windows should trap focus while open and return it when closed. If a visitor cannot complete your contact form or your checkout with a keyboard alone, that is a hard stop, and no widget will fix it from the outside.
Third: form labels and error messages
Forms are where accessibility failures cost actual money. A placeholder is not a label — it vanishes the moment somebody types. Each field needs a real, permanently visible label, and the input should be programmatically associated with it so screen readers announce the right thing.
Errors are the other half. An error message reading 'Invalid input' above a form tells nobody anything. Say which field is wrong and what to do about it, put the message next to the field, and keep whatever the person already typed. This is basic courtesy that happens to be a WCAG requirement.
Fourth: tap targets on mobile
WCAG 2.2 sets a minimum target size of 24 by 24 CSS pixels, and both Apple and Google recommend considerably more than that. Crowded link rows, tiny close buttons on cookie banners and phone numbers styled as four-pixel-high underlined text are all common offenders.
Check it on an actual phone, one-handed, while standing up. If you mis-tap three times in a row, so will a customer with less steady hands than yours.
Fifth: alt text and heading structure
Image descriptions and headings come last not because they matter least, but because they matter most for a minority of visitors and require judgment rather than a simple ratio. Every meaningful image needs alt text describing its purpose; purely decorative images should have empty alt attributes so screen readers skip them. 'Image of team' is not alt text. 'Our fitting team at the Leeds depot, loading a van' is.
Headings should form a logical outline: one H1 per page, H2s for sections, H3s beneath them. Do not pick heading levels for their size. This also happens to be the structure search engines use to understand your page, which is a rare case of the accessibility fix and the SEO fix being the same job.
Why an overlay cannot tick these boxes
Take contrast. An overlay toggling a high-contrast mode changes what one visitor sees in that session; it does nothing about the grey text in your stylesheet. Take labels: if the input has no label element, no script can invent a meaningful one at runtime, because it does not know what the field is for. Take keyboard access: overlays frequently add their own widgets that are themselves difficult to operate without a mouse. Take reading order: a script cannot reliably reorder a DOM that was built as a pile of absolutely positioned divs.
There is also a fair amount of evidence that these tools can interfere with assistive technology people already use, and no regulator anywhere has accepted 'we installed a widget' as evidence of compliance. You would be paying a subscription to look like you have addressed something you have not.
The realistic budget
For a typical small UK service firm, the fundamentals above are usually two to four days of developer time plus some content work. If your site is very old, the form and navigation fixes may cost more, but they are finite, one-off jobs that improve the site permanently.
Compare that with an overlay subscription: recurring, invisible, and attached to code you do not control. The fundamentals win on cost and they win on outcome.
A note on the legal picture
In the UK, the Equality Act 2010 already requires service providers to make reasonable adjustments, and WCAG 2.2 AA is the benchmark everyone points at. The EAA matters directly if you sell into the EU; if you are a UK firm with European customers, it is worth taking seriously rather than assuming Brexit did you a favour. Either way, the practical answer is the same order of operations, done properly, once.
Start with contrast and text size this week. Book the keyboard and form work next. Put alt text and headings into your content process so they stop being a project. And when the next overlay call comes in, ask them which of those five problems their script fixes at source. The pause will tell you everything.