Accessibility and the European Accessibility Act: your shop has until June 2025
From 28 June 2025 the European Accessibility Act applies to online shops. Who is in scope, what WCAG 2.1 AA means for a real checkout, and why an inaccessible shop is losing orders today.
Photo: free stock photography (Unsplash licence) — see imprint
The deadline is real, and it is closer than your roadmap thinks
The European Accessibility Act applies from 28 June 2025. Germany implemented it as the Barrierefreiheitsstärkungsgesetz, and e-commerce is named explicitly among the covered services. That date sounds comfortably far away from a January planning meeting. It is not. If your shop needs template work, a theme rebuild, a checkout audit and a round of retesting, you are looking at a project that competes with everything else on your 2024 list — and everyone else will be booking the same specialists in spring 2025.
We are deliberately not going to quote you a penalty figure. The enforcement details sit with national market surveillance authorities, and anyone who tells you in early 2024 exactly what a fine will look like for a mid-sized German shop is guessing. The honest risk picture is broader than fines anyway: complaints, a market surveillance procedure, an order to fix things on someone else's timetable, and a competitor who did the work quietly pointing it out in a tender. Plan for compliance, not for a number.
Check whether you are in scope before you spend a cent
There is a microenterprise exemption for services, and it is narrower than the people relying on it hope. The threshold is fewer than ten employees together with annual turnover or a balance sheet total not above two million euro. Both conditions, not either. A shop with six employees and three million euro turnover is not a microenterprise. Neither is one that hits nine employees in March and hires two more in May. Get your own numbers confirmed by your Steuerberater rather than by a blog post — including ours.
Note what the exemption does not cover. It applies to service providers, not to products; if you manufacture or import hardware with a digital interface, that is a separate track with its own rules. And in B2B tenders the exemption is worth nothing in practice — public buyers and large corporates have been putting accessibility clauses into procurement for years, and they will not care that the law technically lets you off. If half your revenue comes from framework agreements, treat June 2025 as a sales deadline rather than a legal one.
WCAG 2.1 AA is not a philosophy, it is a list of things your checkout does wrong
The legal texts point at harmonised European standards, which in practice means the WCAG 2.1 AA success criteria. That sounds like a compliance abstraction until you sit down with a keyboard and no mouse and try to buy something from your own shop. Most shops fail within ninety seconds. The cookie banner traps focus. The mega menu opens on hover only. The variant selector is a styled div that Tab skips entirely. The payment step has a custom radio group with no accessible name.
None of that is exotic. It is the accumulated debt of five years of design decisions that were only ever tested with a mouse, on a large screen, by people with good eyesight. The good news is that it is mostly boring, cheap work in the template layer: real form labels instead of placeholder text, visible focus rings your designer removed because they looked untidy, contrast ratios that survive the brand grey on white, error messages tied to the field they belong to. The bad news is there are two hundred of them.
- Keyboard-only checkout: Tab from product page to order confirmation, no mouse, no dead ends.
- Visible focus: every interactive element shows where you are — outline: none is a compliance bug.
- Contrast 4.5:1 for body text, 3:1 for large text and UI borders — measure, do not eyeball.
- Real labels: every input has a programmatically associated label, not a placeholder pretending to be one.
- Errors in text: 'invalid' as a red border only is invisible to a screen reader and to colour-blind buyers.
This is not charity — an inaccessible checkout refuses money
The framing that gets accessibility budgets killed is the framing that treats it as a moral favour to a small minority. Drop it. A blind customer with a screen reader who cannot complete your payment step is not a beneficiary of your goodwill. They are a person who chose your product, entered their address, and then hit a button your developer never gave a name — and they leave and buy the same thing somewhere else. You did not fail to help them. You failed to take their money.
And the group is not small. Permanent disability is only the core of it. Add everyone over sixty reading your 13-pixel grey-on-grey shipping note, everyone with a broken wrist using one hand, everyone in bright sunlight on a phone, everyone on a laptop trackpad they hate. Accessibility fixes are almost always usability fixes wearing a legal hat — larger targets, clearer errors, honest labels, sane focus order. We have never seen those changes lower a conversion rate.
The overlay widget will not save you
Somebody will sell you a JavaScript snippet that promises compliance for a monthly fee and a one-line install. Be extremely sceptical. These tools can adjust font sizes and colours, which is a genuine if modest convenience. They cannot give your variant selector a role, they cannot fix a focus trap in your cookie banner, and they cannot invent alt text that describes what is actually in the photo. Accessibility campaigners have been publicly hostile to overlays for years, and buying one is a good way to be visible and non-compliant at the same time.
Do the work in the templates instead, and sequence it by money. Start with the path that carries revenue: category, product detail, cart, checkout, account. Fix those to a defensible standard, then work outward to the blog and the footer pages. An automated checker such as axe or Lighthouse will find perhaps a third of the problems, which makes it a fine first pass and a terrible sign-off. The remaining two thirds need a human with a keyboard and a screen reader, and that is the part you should budget for now, not in May 2025.
| Area | Typical failure | Effort |
|---|---|---|
| Checkout flow | Custom payment radios with no accessible name | Days — but non-negotiable |
| Forms | Placeholder used instead of a label | Hours, mostly search and replace |
| Colour and contrast | Brand grey on white below 4.5:1 | Hours in CSS, days in brand politics |
| Navigation | Hover-only mega menu, no keyboard path | Often a component rebuild |
| Product images | Alt text is the filename or empty | Cheap per image, brutal at 8,000 SKUs |
- Check the microenterprise thresholds properly — both conditions must hold, and B2B tenders ignore the exemption anyway.
- Fix the revenue path first: category, product, cart, checkout — the footer can wait.
- An automated scan finds roughly a third of the issues; sign-off needs a human with a keyboard.
- An inaccessible checkout is not an ethics problem before it is a revenue problem — it turns away buyers who already decided.
Frequently asked questions
If you sell to consumers online and you are not a microenterprise, assume yes. The microenterprise exemption for services requires fewer than ten employees and turnover or balance sheet total of no more than two million euro — both together, not one or the other. Borderline cases should be checked with your Steuerberater and a lawyer, because headcount and turnover both move.
Honestly: nobody can give you a reliable answer in early 2024, and treat anyone who quotes a precise fine with suspicion. Enforcement runs through national market surveillance authorities and starts with complaints and procedures rather than instant penalties. The practical risk is being told to fix things on someone else's schedule, plus losing B2B tenders that require accessibility regardless of the law.
No. Overlays can offer font and colour adjustments, which is a small real benefit, but they cannot repair semantics they did not author: unlabelled controls, focus traps, hover-only menus, missing alt text. Buying one does not remove the underlying failures, it just adds a monthly fee on top of them. Spend the same money on template work and a keyboard test instead.
For a shop on a close-to-standard theme, an audit plus the checkout and forms work is typically a few weeks of focused effort. A heavily customised theme with bespoke sliders, filters and a custom checkout can take several months, because those components usually need rebuilding rather than patching. Start with an audit — it is cheap and it tells you which of those two projects you actually have.
The legal picture for pure B2B is less clear-cut than for consumer commerce, and that is a question for your lawyer, not for us. Commercially it is not close: your buyers are employees, some of them use assistive technology, and public sector and corporate procurement have been putting accessibility requirements into tender documents for years. A B2B portal that fails a keyboard test loses deals long before it loses a legal argument.
We do this for a living — Shopware, Node.js, React, ERP integration and automation for B2B.
Talk to an engineer