Website Accessibility Audit: DIY Checklist, Automated Scanner, or Manual Audit?

A website accessibility audit is the process of checking a site against WCAG 2.1 (or 2.2) success criteria to find barriers that block people with disabilities. You can run one three ways: a manual DIY checklist, an automated scanner, or a paid manual audit by an accessibility specialist — and most sites actually need a mix of the first two, with the third reserved for legal risk or complex apps.

I get asked some version of "which one do I actually need" almost every week, usually from someone who just got a demand letter or is trying to avoid getting one. So let's break down what each method actually catches, what it costs, and where the gaps are — because there are gaps in all three, and pretending otherwise is how sites end up with a false sense of security.

What's the difference between a DIY checklist, an automated scanner, and a manual audit?

A DIY checklist is a self-guided review using a published list of WCAG rules — you go page by page and check things off by eye. An automated scanner crawls your site's code and flags violations programmatically, usually in seconds per page. A manual audit is a paid engagement where a trained auditor (often using a screen reader themselves) tests your site by hand against the full WCAG standard, including things software literally cannot judge.

Here's the part nobody likes to hear: none of these three replace each other. They catch different things, at different depths, for different costs. Treating any single one as "the audit" is how real barriers slip through.

Quick comparison

Method	Typical cost	Coverage	Best for
DIY checklist	Free (your time)	~20-30% of issues, surface-level	Small sites, early awareness
Automated scanner	Free to $49/mo	~30-40% of WCAG success criteria	Ongoing monitoring, catching regressions
Manual audit	$1,500-$25,000+ per site	~90-100% when done well	Legal risk, complex apps, VPATs

Those coverage numbers line up with what Deque and other accessibility research groups have reported for years: automated tools reliably catch roughly a third of WCAG issues, because the rest require human judgment — context, meaning, and intent that code alone can't evaluate.

What can a DIY accessibility checklist actually catch?

A DIY checklist catches the obvious, visible stuff: missing alt text, low color contrast, form fields without labels, and pages missing a page title. It's a decent gut check for someone with zero budget, but it depends entirely on the reviewer knowing what to look for — and most marketers and site owners don't.

I've watched a small business owner run through a printed WCAG checklist for two hours and miss a completely broken tab order, because you can't see tab order. You have to unplug your mouse and try it. That's the core weakness of DIY review: it's a visual pass on a standard that's mostly about non-visual experience.

Where a checklist genuinely helps: heading structure, image alt text, link text that makes sense out of context, and basic color contrast (a free tool like WebAIM's contrast checker handles that part in seconds). If you want the full rule list to work from, our WCAG checklist covering every rule is written in plain English specifically so non-developers can use it.

What it almost never catches: keyboard traps, screen reader announcement issues, focus order problems, and anything that depends on how assistive technology actually parses your DOM. That stuff needs either software that reads the code, or a human who uses a screen reader daily.

How good is automated accessibility testing, really?

Automated accessibility testing is fast, cheap, and consistent — but it maxes out at somewhere between a third and 40% of WCAG success criteria, because plenty of rules require human interpretation. A scanner can tell you an <img> tag has no alt attribute. It cannot tell you whether the alt text you wrote is actually accurate or just keyword-stuffed junk.

That said — automated testing is where most sites should start, and honestly, where most sites should live permanently for maintenance. Here's why: accessibility isn't a one-time fix. Every new blog post, every new form, every redesigned nav bar can reintroduce old bugs. A scanner that runs weekly catches regressions a one-time manual audit never will, simply because the manual audit happened eight months ago and your dev team shipped eleven releases since then.

The WebAIM Million study, which scans the homepages of the top one million sites every year, consistently finds an average of 50+ accessibility errors per homepage — and that's just what automated detection can find. Real-world numbers are almost certainly higher once you add human testing into the mix.

What automated scanning is genuinely good at:

• Missing or empty alt attributes on images
• Color contrast ratios below the 4.5:1 (or 3:1 for large text) threshold
• Missing form labels and empty buttons or links
• Heading structure problems — skipped levels, multiple <h1> tags, missing headings entirely
• Missing lang attributes, missing document titles, and duplicate IDs

What it can't reliably judge: whether alt text is meaningful, whether a custom widget announces state changes to screen readers correctly, or whether your site's reading order actually makes sense when tabbed through. If you're not sure what heading mistakes look like in practice, we wrote a deep dive on heading structure mistakes that break both WCAG compliance and AI extraction — because the same sloppy markup that confuses a screen reader also confuses ChatGPT and Perplexity when they try to parse your page.

This is actually where automated tools have gotten more useful over the last two years — not just for accessibility, but for a second, related problem: whether AI search engines can even read your content in the first place. That's a big enough shift that it deserves its own explanation, which is why we put together a full guide on what AI readability actually means and how to check yours.

How much does a manual accessibility audit cost?

A manual accessibility audit typically costs anywhere from $1,500 for a small marketing site to $25,000 or more for a large e-commerce platform or web app, depending on page count, complexity, and whether the auditor tests with real assistive technology users. Enterprise audits that include a full VPAT (Voluntary Product Accessibility Template) for procurement purposes often land at the higher end of that range, sometimes higher still.

Pricing usually breaks down by a few factors:

• Page/template count — auditors typically test a representative sample of unique templates, not every single page, which is how a 400-page site can still get audited in a reasonable timeframe.
• Depth of testing — a basic manual review with a checklist and one screen reader pass costs less than a full audit that includes testing across NVDA, JAWS, VoiceOver, and keyboard-only navigation.
• Deliverables — a simple issue list costs less than a full VPAT with legal-ready documentation, which some enterprise buyers and government contractors specifically require.
• Remediation support — some firms quote the audit separately from the fix work; others bundle it. Ask which one you're getting, because the fix work is usually 3-5x the audit cost.

Is it worth it? For a site with real ADA litigation exposure — think retail, healthcare, education, or hospitality with a lot of consumer traffic — yes, almost certainly. Manual audits catch the nuanced stuff: whether your custom date picker actually works for a screen reader user, whether your video captions are synced correctly, whether a modal dialog traps focus properly. Software just doesn't judge that reliably yet.

For a five-page local service business site? A manual audit is probably overkill. Start with automated scanning, fix what it finds, and revisit the manual audit question if you scale up or start seeing legal risk (a demand letter, a state AG inquiry, that kind of thing). If you want the legal backdrop on why any of this matters in the US, our piece on how ADA and WCAG actually relate to each other covers what compliance legally requires versus what's just best practice.

Which method should you actually use?

Use automated scanning as your baseline, layer in a DIY manual check for the things software can't judge, and bring in a professional manual audit only when your risk profile or site complexity actually justifies the cost. That's the honest answer, and it's not the tidy "pick one" answer people usually want.

Here's a rough decision framework I'd actually use:

• Under 20 pages, low traffic, tight budget: automated scanning plus a manual DIY pass on your top 5 pages. Skip the paid audit for now.
• Growing site, e-commerce, lead gen forms everywhere: automated scanning running weekly, plus one manual audit annually to catch what the scanner misses.
• Enterprise, healthcare, government, or you've already gotten a demand letter: full manual audit with a VPAT, then automated monitoring in between audits so you're not paying for a full re-audit every quarter.

One thing I'll push back on directly: overlay widgets that promise "instant compliance" are not a substitute for any of the three methods above. They patch the surface with a JavaScript layer and leave the underlying code broken — which is exactly what both automated scanners and manual auditors are trained to detect. We've written at length about why overlay widgets fail WCAG audits, and honestly, if a vendor is selling you a one-line script as a full fix, that's a red flag worth taking seriously.

Does an accessibility audit help with AI search visibility too?

Yes — and this is the part most audit checklists still leave out entirely. The same code-level issues that block screen readers (bad heading structure, missing alt text, unlabeled buttons, poor semantic markup) also make it harder for AI crawlers like GPTBot, PerplexityBot, and Google's AI Overview crawler to parse and cite your content accurately.

Semantic HTML — using a <nav> element for navigation instead of a generic <div>, using real <button> elements instead of clickable spans — tells both a screen reader and an AI crawler what a piece of content actually is. Skip it, and both audiences lose. We go deeper on that overlap in our post on why semantic HTML improves accessibility and AI visibility at the same time.

This is exactly the gap AccessKnight was built to close — it runs both checks in one pass, scanning against all 33 WCAG rules and scoring AI readability at the same time, so you're not running two separate audits for two separate problems that turn out to share the same root causes.

Frequently Asked Questions

How much does an automated accessibility scan cost?

Automated scans range from free to around $50/month for ongoing monitoring. Free tiers usually cap the number of URL scans per month (AccessKnight's free plan, for example, allows 5 scans a month), while paid tiers add unlimited scanning, scheduled monitoring, and multi-site tracking for agencies.

Is a DIY accessibility checklist enough to avoid an ADA lawsuit?

No. A DIY checklist reduces risk but doesn't eliminate it, because it depends on the reviewer's knowledge and misses issues that require assistive technology testing, like keyboard traps or screen reader announcement bugs. Most ADA demand letters cite specific, testable WCAG failures — the kind automated scanning and manual audits are built to catch, but a casual visual review often isn't.

How long does a manual accessibility audit take?

A manual audit typically takes 1-4 weeks depending on site size, ranging from a few days for a small marketing site to several weeks for a large web application with multiple user flows and templates. Timeline usually scales with page count and how many assistive technologies the auditor tests against.

Can automated tools replace a manual accessibility audit entirely?

No. Automated tools typically catch around 30-40% of WCAG success criteria, leaving issues like logical reading order, meaningful alt text, and correct ARIA usage — things that require human judgment — undetected. The two methods are complementary, not interchangeable.

What's the difference between WCAG and ADA compliance?

WCAG (Web Content Accessibility Guidelines) is the technical standard; ADA (Americans with Disabilities Act) is the US law that courts have interpreted as requiring websites to be accessible, generally using WCAG as the benchmark. There's no official government-mandated WCAG version in the ADA text itself, but WCAG 2.1 AA is the de facto standard cited in most settlements and consent decrees.

How often should I re-run an accessibility audit?

Automated scans should run weekly or after every major site update, since new content and code changes can reintroduce old issues. A full manual audit is typically worth repeating annually, or whenever you do a significant redesign or add new interactive features.

Next step

If you haven't run any kind of audit yet, don't start by pricing out a $10,000 manual engagement — start by seeing what's actually broken. Run a free scan at accessknight.com, get a code-level list of what's failing, and use that to decide whether you need a DIY fix session this weekend or a bigger conversation with a specialist.