Private services under the EAA
Specified consumer services — e-commerce, consumer banking, certain transport, electronic communications, e-books — from 28 June 2025, subject to scope and transitions.
Primary source: Directive (EU) 2019/882WCAG 2.1 AA · EN 301 549 · scheduled
Know where your site stands.Prove what you fixed.
Reachcanon scans the site you are responsible for on a schedule, names the rule each element breaks, maps it to WCAG and EN 301 549, and keeps a dated record that still answers questions months later. It does not certify compliance — no scanner can.
No card. No sales call. One verified site, a monthly scan and the full evidence report on the free plan.
- WCAG 2.1 A/AA criteria tested
- 24/50
- automated checks in the live rule set
- 72
- PASS outcomes — by design
- 0
reachcanon.com/sites/shop.exampleIllustrative example
shop.example
Weekly
4open now
down from 14
Open issues, last eight weekly scans
| Open issues, last eight weekly scans | Open issues | No longer detected |
|---|---|---|
| W1 | 14 | 0 |
| W2 | 12 | 2 |
| W3 | 11 | 4 |
| W4 | 9 | 6 |
| W5 | 11 | 6 |
| W6 | 7 | 9 |
| W7 | 6 | 11 |
| W8 | 4 | 13 |
2 issues came back
Scan of 14 Oct · after a deploy. Alert sent to the owner.
Contrast 3.45:1 against #ffffff — 4.5:1 expected
a.promo-link · /products/rain-shell
- axe rule: color-contrast
- WCAG SC: 1.4.3
- EN clause: 9.1.4.3
Claims you can check before you sign up
Nothing on this page asks to be taken on trust.
01The chain
Every finding names the axe rule, the WCAG success criterion and the EN 301 549 clause.
See the mapping02The boundary
What we test, what we cannot, and what we have not built yet are three separate lists, published.
Read the coverage table03No PASS
A rule that did not fire reports “no issue detected”. Undecidable checks report “needs review”, never a pass.
Read the method04Our own statement
Reachcanon publishes its own accessibility statement, including what is not yet right.
Read our statement
Why anyone buys this
Three questions you probably cannot answer right now.
Not “is my site accessible” — that one has no useful answer. These are the three that come up when a complaint lands, when a public-sector client sends a procurement questionnaire, or when your own team asks what to do first.
Where exactly does it fail?
Your developers say it is fine. A consultant sends a 40-page PDF. Neither gives you a list of elements with the rule each one breaks, which is the only form the work can actually be scheduled from.
Can you prove what you fixed, and when?
Accessibility work is continuous, so the question is never “is it fixed” but “what did you know, and when”. Without a dated record per element, the honest answer is that you cannot say.
What accessibility information should you publish?
Public-sector accessibility statements and the information duties for covered private services come from different rules. Where a statement is appropriate, start from your findings and complete the organisation-specific information. Review the applicable national requirements before publishing.
How it works
From a URL to a record that holds up, in six steps.
The same loop runs for one site or fifty. Each step is a screen in the product, not a service you book.
1. Add the site
Paste the address. Group sites by client if you are an agency.
2. Verify ownership
A file or a DNS record proves you control it. A one-off scan of public pages needs none; scheduled monitoring and signed-in scans do.
3. First scan
A real browser crawls and tests each page. The first scan sets the baseline.
4. Triage by impact
Issues sorted by severity and criterion. Assign, mark fixed, or accept the risk with a reason.
5. Watch for regressions
Scheduled scans compare with the last one; on Professional and Portfolio the team is emailed when something comes back.
6. Report and draft
PDF evidence report and CSV/JSON exports on every plan; on Professional and Portfolio, a draft statement where one applies.
One finding, in full
Written to be read by somebody hostile, months later.
Findings are append-only. A later scan writes new rows rather than editing old ones, and your response to a finding is recorded beside it — so the timeline can be reconstructed instead of taken on trust.
Every scan exports as CSV or JSON with the disclosures inside the file, because a caveat that lives only on a dashboard is missing from the document somebody forwards to a regulator.
FailureSerious1.4.3 Contrast (Minimum)
Contrast 3.45:1 against #ffffff. Expected 4.5:1.
a.promo-link · /products/rain-shell
As published#8a8a8a on #ffffffFails AA3.45:1
Corrected#767676 on #ffffffPasses AA4.54:1
4.5:1 — the minimum WCAG 2.1 AA allows for body text
Every finding carries its whole chain
- color-contrastaxe-core rule
- 1.4.3 Contrast (Minimum)WCAG 2.1 AA
- 9.1.4.3EN 301 549
Observed 26 September 2026, 04:12 UTC · axe-core 4.13.0
01
The element
The selector and a snippet, precise enough to find it again after a redesign.
02
The rule and the clause
What fired, the WCAG criterion, and the EN 301 549 clause a regulator indexes by.
03
The moment
When it was observed and the exact engine version — two scans can differ because the engine improved, not because you regressed.
04
Your response
Fixed, will not fix, third-party component — with the reason, which is the part that actually gets asked about.
Why it is a schedule, not a one-off
One scan tells you about today. A schedule tells you whether it stayed fixed.
Your site ships every week. A contrast fix made in March gets undone by a redesign in June, and a report from March says nothing about it. That is why Reachcanon is priced on how often a site is watched rather than on a count of scans.
Open failures, eight weekly scans
Catch the regression, not the anniversary
A weekly scan finds the failure your last deploy introduced while the person who wrote it still remembers the change.
A trend is an answer; a snapshot is a claim
Asked what you are doing about a failure, a line showing it closing over six months answers the question. One clean scan from last year does not.
Evidence that survives staff turnover
The record outlives whoever ran it: page, element, rule, clause, timestamp, engine version, and the reason your team recorded for anything left unfixed.
The part most tools leave out
A clean scan is not conformance, and this one says so on every report.
Automated testing reaches roughly a third of accessibility failures. A tool that shows you a green score for the other two thirds is selling reassurance — and reassurance is exactly what fails under scrutiny.
What a scan can decide
- Contrast ratios it can compute, missing alternative text, unlabelled form controls.
- Landmark structure, heading order, accessible names on interactive elements.
- Page language, a visible focus indicator on each control, reflow at 320px wide, and ARIA attributes that reference nothing.
What only a person can judge
- Whether the alternative text actually describes the image.
- Whether captions are accurate and synchronised with the audio.
- Whether your support desk can serve a disabled customer at all.
So there is no PASS in this product. A rule that ran and did not fire reports as “no issue detected”, which is a smaller and truer claim. A check the engine could not decide reports as “needs review” and is never folded into a percentage to make the number look better.
And it can be wrong. A rule engine can miss a real failure, and it can flag something a person would judge acceptable in context. Findings are where your review starts, not a complete list — and pages the crawler could not reach are unassessed, not compliant.
No score. No certificate. No badge for your footer.
Published before you pay, not discovered after
These parts of EN 301 549 cannot be judged by scanning a page. They are listed here, on the marketing site, because finding out after purchase is how a tool loses your trust.
| EN 5 | Generic requirements | Hardware, closed functionality and biometrics — not a web page. |
| EN 6 | Two-way voice communication | Real-time text and caller ID behaviour cannot be observed by a page scan. |
| EN 7 | Videos with audio | Whether captions are accurate and synchronised is a human judgement. |
| EN 11 | Software | Native applications outside the browser are not in a web crawl's reach. |
| EN 12 | Documentation and support services | Whether your help desk can serve a disabled customer is an operational fact, not a technical one. |
| EN 13 | ICT providing relay or emergency service access | Out of scope for a website scan entirely. |
Who it is for — and who it is not
The European Accessibility Act does not cover every website.
We would rather you knew this before signing up. Whether a duty applies to you is a legal question about your service; a scan cannot answer it and Reachcanon never claims to.
Microenterprises providing services
Fewer than 10 staff and no more than €2 million turnover or balance sheet: exempt from the EAA's service requirements. Accessibility is still worth doing; it is not this law's duty.
Primary source: Directive (EU) 2019/882Public-sector bodies
A separate regime, the Web Accessibility Directive (EU) 2016/2102, which governs public-sector websites and their accessibility statements.
Primary source: Directive (EU) 2016/2102Not legal advice. Review the applicable national rules, or ask counsel, before relying on any scope decision.
A draft statement, where applicable
A statement that names the failures, drafted from your own scan.
Reachcanon drafts it from your latest scan: the criteria that failed, the checks still under review, and a disproportionate-burden paragraph carrying the reason your team recorded. Review your applicable statement or information duties, complete what a scan cannot know, and approve the draft before publishing.
It is a draft, and it stays a draft until you publish it. It is a declaration about your organisation, not about ours.
# Accessibility statement
Parts of this website are not fully accessible. Automated testing on 26 September 2026 found failures against WCAG 2.1 AA success criteria 1.1.1, 1.3.1, 1.4.3, 2.4.4, 3.1.1 and 3.3.2.
## Disproportionate burden
2.5.3 Label in Name: the quantity stepper is supplied by our basket vendor and cannot be restyled without forking their component.
## Preparation of this statement
It is based on automated testing, which reaches approximately one third of accessibility failures. No independent auditor has reviewed this website.
What lands after a scan
You get a file, not a dashboard you have to screenshot.
Everything below is produced from the same scan record, so the PDF, the spreadsheet and the published statement cannot disagree with each other or with what the console shows.
01
A findings list you can schedule work from
Grouped by success criterion rather than dumped as rows — each one naming the pages it affects, the element it is on, and what the check objected to.
02
A print-ready PDF evidence report
A4, paginated, with the criteria, the coverage boundary, the pages tested and the method. Designed to be attached to a procurement response or handed to a regulator, not to be read on a phone.
03
The raw evidence, as CSV and JSON
Every finding with its page, element, rule, criterion, EN clause and the moment it was observed. The disclosures travel inside the file, because a caveat left on a dashboard is missing from the document somebody forwards.
04
A drafted accessibility statement
Written from your latest scan in English or Spanish, naming the criteria that failed and carrying the disproportionate-burden reasons your team recorded. A draft you correct and publish, never one we publish for you.
05
A record that survives the people who made it
Findings are append-only and your responses sit beside them, so a year later the question “what did you know, and when” has an answer instead of an argument.
06
What changed since the last scan
Each scan is compared with the one before it, so a regression a redesign introduced shows up as a new failure rather than disappearing into a total. This is the whole argument for a schedule: one scan tells you about today.
Pricing
Priced on how often your site is watched.
Every plan states its ceiling. Nothing here is unlimited — this product drives a real browser against real sites, so an unbounded plan would be a promise the infrastructure cannot keep.
Free
Scanned monthly
€0per month
Verify one site and keep the report. No card, no expiry.
What it buys
See what fails, element by element, and keep the PDF evidence report. No card.
- 1
- site
- 1
- scheduled scan a month
- 10
- pages per scan
- 1
- extra scan you can run by hand
- 3
- months of history retained
- Evidence export: PDF report, CSV and JSON
- Statement drafting
- Public pages only
- Regression alerts need weekly or daily
Best value for one site
Starter
Scanned monthly
€29per month, excl. VAT
€29 per watched site a month
One site, checked every month so you know where you stand.
What it buys
Five times deeper than free — 50 pages a scan, 8 extra scans on demand and a year of dated history.
- 1
- site
- 1
- scheduled scan a month
- 50
- pages per scan
- 8
- extra scans you can run by hand
- 12
- months of history retained
- Evidence export: PDF report, CSV and JSON
- Statement drafting
- Public pages only
- Regression alerts need weekly or daily
Professional
Scanned weekly
€99per month, excl. VAT
€33 per watched site a month
You ship weekly and need to catch a regression in the same week.
What it buys
Catch a regression the week it ships: 3 sites scanned weekly, email alerts, signed-in pages tested, statements drafted, 3 years of history.
- 3
- sites
- 12
- scheduled scans a month
- 150
- pages per scan
- 40
- extra scans you can run by hand
- 36
- months of history retained
Page-level detail for the latest 12 months; monthly summaries before that.
- Evidence export: PDF report, CSV and JSON
- Statement drafting
- Scans behind your sign-in (cookie or header)
- Regression alerts
Portfolio
Scanned daily
€299per month, excl. VAT
€20 per watched site a month
An agency or a group monitoring many sites from one workspace.
What it buys
Every client site checked daily: 15 sites, 400 pages a scan, alerts and statements, and six years of history, page by page for the latest year.
- 15
- sites
- 450
- scheduled scans a month
- 400
- pages per scan
- 50
- extra scans you can run by hand
- 72
- months of history retained
Page-level detail for the latest 12 months; monthly summaries before that.
- Evidence export: PDF report, CSV and JSON
- Statement drafting
- Scans behind your sign-in (cookie or header)
- Regression alerts
Prices exclude VAT. Billing, cancellation and refunds: terms of service. Personal data you put in Reachcanon: data processing agreement.
Paid plans open soon
The free plan works in full today: one verified site, a monthly scan and the complete evidence report, with no card. Paid plans are listed so you can see every limit before you commit; they go on sale once card payments open, and the evidence you build on the free plan carries over.
Scan the site you are responsible for.
You get findings mapped to the standard, the parts no scan can judge listed rather than hidden, and a record that holds up later. No score.
Request an invitation