Start your 7-day free trial, card not charged until it ends
Browse all topics

Accessibility monitoring

Brand, compliance, accessibility & SEO · Last updated 16 July 2026

Roughly one in five people has a disability, and many of them rely on a screen reader, keyboard navigation or high-contrast settings to use the web. If your site is hard for them, they leave, and in many places an inaccessible site is also a legal risk. Accessibility scanning audits your pages against the WCAG 2.2 AA standard, gives each scan a grade, and watches for regressions.

What it is

An automated audit that loads your pages, runs the industry-standard axe-core accessibility engine over them, then adds an AI reviewer for issues automated rules miss. It scores each scan against WCAG 2.2 AA, gives it a grade from A to F, and lists every issue with the element and the guideline it breaks.
Read more
axe-core is the same open-source engine that powers many professional accessibility tools; it deterministically catches issues such as missing image alternatives, poor colour contrast, unlabelled form fields and broken landmark structure. The AI reviewer then looks for the more judgement-based problems (confusing link text, questionable reading order) and is told not to repeat the issues axe-core already checks for. Each finding records the WCAG success criterion it relates to, the element selector, and where on the site it appears.

Why it matters

Accessibility problems are invisible to most of the people who build a site, because they are not using a screen reader or navigating by keyboard. They stay hidden until a real user hits a wall, or a complaint arrives. An automated audit surfaces them while they are cheap to fix.
Read more
Accessibility is both the right thing to do and, increasingly, a legal requirement: public-sector bodies and many businesses are held to WCAG AA by law. But it is also a moving target, because every content change can introduce a new barrier. Auditing on a schedule, with a grade you can watch, is how you keep a site accessible instead of fixing it once and letting it drift.

This is detection, not certification

Automated testing, even axe-core plus an AI reviewer, catches only part of what makes a site accessible. Many WCAG criteria can only be judged by a human actually using assistive technology. This audit is a strong, continuous first line of defence that finds the issues machines can find and keeps you honest between manual reviews. It is not a legal accessibility certification or a substitute for testing with real users. Treat a clean grade as "no machine-detectable barriers", not "guaranteed compliant".

The grade and score

Every scan produces a numeric score out of 100 and a letter grade. The score starts at 100 and subtracts a weight for each issue found, according to how serious it is:

  • A critical issue subtracts 25.
  • A serious issue subtracts 10.
  • A moderate issue subtracts 3.
  • A minor issue subtracts 1.

The score then maps to a grade: 90 or above is A, 80+ is B, 70+ is C, 55+ is D, 40+ is E, and below 40 is F. These are the same severity bands (critical, serious, moderate, minor) that axe-core itself uses, so the grade reflects real-world impact, not just a count.

What each finding shows

Every issue in a report tells you what is wrong, where, and why it matters:

  • The message and a fix. What broke, in plain terms, a "How to fix" note and, for axe-core findings, a link to a full write-up on the rule.
  • The WCAG success criterion it relates to (for example 1.1.1 for missing text alternatives, or 1.4.11 for contrast), so you can tie the fix back to the standard.
  • The impact level: critical, serious, moderate or minor.
  • The element (a CSS selector and a snippet of the offending HTML) and which pages it appears on, since one template issue often repeats across the site.
  • Whether it came from axe-core or the AI reviewer, labelled "Automated" or "AI-assisted".

Baseline and continuous monitoring

You can run a one-off audit any time, but the real value is watching for regressions. Turn monitoring on for a domain and TLDTrack re-scans it on a schedule and compares each scan against a baseline:

  • The first monitored scan seeds the baseline: the set of issues that exist today, which you have chosen to accept for now. It does not alert.
  • Each later scan is compared against that baseline, so an alert is about new issues introduced since, not the backlog you already knew about.
  • You can re-set the baseline at any time to "bank" fixes you have made, so future alerts are measured against the improved state.
  • Choose a cadence of daily, weekly (the default) or monthly.
  • Scheduled monitoring scans check the homepage. To audit more pages, run a scan yourself.

The grade floor

Alongside new-issue detection, you can set a grade floor: the lowest grade you are willing to tolerate. If a monitored scan comes back worse than your floor (say you set B and the site drops to C), that triggers an alert on its own, even if no brand-new issue was introduced. It is a simple way to say "tell me the moment this site slips below the standard I hold it to."

What you need before you start

  • A monitored website to audit. Add the site first if you have not (see Add your first website).
  • Scan credits. Accessibility draws on the shared scan-credit pool, at one credit per ten pages, rounded up.
  • A sense of which pages matter. You can discover your pages and select up to 50 to audit when you run a scan yourself; scheduled monitoring checks the homepage.

How alerting works

A completed monitoring scan alerts the account owner and your notification contacts when either of two things is true: at least one new issue has appeared since your baseline, or the grade has dropped below your floor. The email lists the top issues, worst first, and push notifications tell you how many new issues need review or that the grade has slipped. A scan that introduces nothing new and stays above your floor does not alert, and scans you run yourself never alert, so you hear about regressions, not routine passes.

How to set it up, step by step

  1. 1
    Open the site from your Domains list and go to its Accessibility tab.
  2. 2
    Discover the site's pages and select the ones to cover (up to 50), or run a quick audit of the key pages.
  3. 3
    Run a scan. It loads each page, runs axe-core and the AI reviewer, and returns a score, a grade and the full list of issues within a few minutes.
  4. 4
    Work through the issues worst first. Critical and serious ones (missing labels, poor contrast, keyboard traps) are where to start.
  5. 5
    When you are ready to monitor, switch on Monitor continuously, pick a cadence (daily, weekly or monthly), and set a grade floor if you want to be told when the site slips.
  6. 6
    Use Set current as baseline whenever you want future alerts measured against the current, fixed-up state.

Fix templates, not pages

Most accessibility issues live in a shared template or component, not on one page, which is why a single finding often lists many pages. Fixing the template clears the issue everywhere at once and lifts the grade across the whole site, so prioritise the findings that repeat.

Where to set this up

Good to know

  • Accessibility shares the scan-credit pool with security scans, compliance and SEO audits. One credit covers up to ten pages.
  • The audit targets WCAG 2.2 AA, the level most laws and policies reference.
  • Baseline comparison means alerts are about new problems, not the backlog you already know about.
  • The grade floor lets a slip below your chosen standard alert you on its own.
  • Automated testing is a first line of defence, not a certificate. Pair it with real manual testing for anything you must legally certify.
Available on:StarterProAgency

Was this guide helpful?

If anything is unclear or missing, tell us and we will improve it.

Contact support