How to tell whether an accessibility audit is worth paying for
What an automated scan report is really telling you
"Two quotes for the same audit came back looking nothing alike, and now I don't know which one is real." That's usually where this starts. One proposal is a page and a half. The other runs forty pages with screenshots. Before comparing prices, it helps to know what each kind of report is actually built from.
A scan report is generated by software that crawls your pages and checks them against a rule set. Missing alt text. Insufficient contrast. Form fields with no label attached. It runs in minutes and it's often free or nearly free.
It's genuinely useful as a first pass, and it will catch real problems worth fixing. What it can't tell you is whether a page makes sense to someone actually using a keyboard or a screen reader. A scanner has no way to judge whether the experience holds together as a whole.
A scan report often runs to dozens of flagged items on a site that's never been touched for accessibility before. That length can feel alarming. Most of those items are small, individually fixable things, not evidence the site is a disaster.
What manual testing adds that a scan cannot produce
Manual testing means a person, sometimes using the same assistive technology a visitor would use, actually going through your site the way that visitor would. Not a checklist run by software. A real attempt to complete a real task.
They notice things a scanner structurally cannot. A form that passes every automated check but still confuses someone navigating by keyboard. An image whose alt text is technically present but says nothing useful. A menu that opens with a mouse click and nothing else.
This is the same distinction behind testing your own site with a keyboard before paying anyone. A scan and a human pass surface genuinely different classes of problem. A serious audit includes both, not one standing in for the other.
Ask whether the person doing the manual pass actually uses assistive technology day to day, or whether they're a developer running through a checklist for the first time. Both can produce a useful report. They are not the same depth of testing, and the price should reflect which one you're getting.
Reading the proposal: pages in scope, sampling, retest
When two audit quotes look nothing alike, the proposal is where the real difference lives. Not the number at the bottom of the page.
Ask how many pages are actually in scope. Is that every template on the site, or a sample of a few? Ask whether the testing is manual, automated, or both. Ask whether the report includes a retest once fixes go in, or whether that's billed separately.
A proposal that answers all of that in plain language, before you sign anything, is telling you something about how the rest of the engagement will go. One that dodges the question is telling you something too, and it's worth listening to that as much as the answer itself.
An audit with no remediation attached is a document you cannot use
An audit that hands you a list of problems with no guidance on fixing them is a document you can read but not act on. Somebody still has to translate "insufficient color contrast on primary buttons" into an actual change to a stylesheet. That translation is where most of the real value sits, not in the list of problems itself.
Ask directly, before you pay: does the audit include remediation guidance specific enough for a developer to implement without guessing? Some audits include a short call to walk through the findings together. Others hand over a spreadsheet and consider the job done. Know which one you're buying before the invoice arrives.
A short walkthrough call, even thirty minutes, tends to be worth more than an extra ten pages in the written report. Questions come up the moment someone starts trying to act on a finding, and a live conversation answers them faster than another round of email.
Sites where an audit is the wrong first purchase
If your site is brand new, still being built, or scheduled for a redesign soon, an audit of the current version can mean paying to examine something you're about to replace. That money is better spent elsewhere, at least for now.
The better order, in that case, is to ask the person building the new site about accessibility before you sign, so the next version gets built right instead of audited and fixed twice.
An audit earns its keep on a site that's staying roughly as it is for a while. A written report gives you something concrete to hand a developer and work through over time, one item at a time.
A site mid-redesign is the clearest exception, but a site with a rebuild vaguely on the someday list is not the same thing. If there's no funded plan and no date, the current site is what visitors are using right now, and it's worth treating that way.
Who should own the fix list once the report lands
A finished audit report sitting in an inbox fixes nothing on its own. Somebody has to actually do the work it describes.
Decide, before the report arrives, who's responsible for working through it. The person who built the site. Someone new. Whoever normally handles updates day to day. That's also where vetting who builds or fixes a site in general becomes relevant, since not everyone who can build a page can also read and act on an accessibility report competently.
One more thing worth asking before any of this: whether an overlay widget was pitched to you as a shortcut around all of it. Whether those widgets actually work is a fair question to settle first, connected to the reasoning behind accessibility work on a small site.
More from the blog
- Will a professionally built website actually get you more customers
- When a booking system is worth adding to your site
- What has to be on your website before a stranger will choose you
- What to ask before a monthly retainer adds a new tool to your site
- Where your website's contact form submissions actually go