If you need to persuade somebody who controls a budget, you need numbers that survive being checked. Plenty of accessibility marketing uses figures that do not. So here is the case, with each claim tied to a source you can open, and with the claims we refuse to make marked as such.
How Many People This Affects
The World Health Organization estimates 1.3 billion people, about 16% of the world, or one in six of us, live with a significant disability. That is not a rounding error.
Narrowing to the ones your website actually affects:
- Vision. WHO puts the number of people with a near or distance vision impairment at at least 2.2 billion globally, of whom around 1 billion have an impairment that could have been prevented or is still unaddressed.
- Colour blindness. The National Eye Institute puts it at about 1 in 12 men, a figure usually quoted for men of northern European descent, and Colour Blind Awareness adds 1 in 200 women. On a site with 10,000 monthly visitors that is several hundred people, and what it means in practice is that red and green are hard to tell apart, so a red border round a field says nothing to them.
- Motor and dexterity. Tremor, arthritis, repetitive strain and temporary injury all push people onto the keyboard.
- Cognitive and learning. The area least designed for, and the standards body says so about its own standard. WCAG's introduction concedes that content conforming at the highest level still will not be accessible to everybody, and names the cognitive, language and learning areas as where the gap is widest. It explains why, too. The working group could not reach consensus on how to make more of it testable. Plain language, predictable layouts and clear errors are most of the intervention.
Then add the temporary and the situational. A broken wrist, an eye infection, sunlight on a phone screen, a train with no headphones. Accessible design is what holds up in all of those, which is why the benefit is much wider than the disability statistics alone suggest.
One economic figure is worth having, because it comes from a UN body rather than from a vendor. WHO estimates the annual global productivity loss from vision impairment at around US$411 billion, against a US$25 billion gap in what it would cost to meet the unmet vision care need. That is not a claim about your revenue and it is the right order of magnitude for what exclusion costs.
What It Costs to Ignore
The legal exposure is real and measurable. 3,117 website accessibility lawsuits were filed in US federal court in 2025, up 27% on 2024, making up 36% of all ADA Title III filings. Counting state courts, filings exceeded 5,000. Our lawsuits page has the detail and the sources.
But litigation is the expensive, late signal. The earlier and quieter cost is the one nobody measures: people who could not finish and left. A checkout that cannot be completed by keyboard does not generate a complaint. It generates an abandoned cart. It looks exactly like every other abandoned cart in your analytics.
The claim we will not make
You will see very large figures quoted for the "spending power of the disability market", usually attributed vaguely. Those numbers mix household income, global populations, and non-digital spending, and they do not survive scrutiny. We are not going to use them. The honest version is simpler. Some proportion of your visitors cannot complete tasks on your site today, and you can find out how many by testing.
The Million-Page Scan, and the Year the Trend Turned
WebAIM scans the top million home pages every year, and the 2026 results are the first bad ones in six years. 95.9% of them carried detectable WCAG failures, up from 94.8%, reversing a run of small improvements in each of the previous six years. That means the web has not been quietly getting better while everybody was busy, and the direction of travel has just changed to the wrong one.
The averages moved with it. The typical home page carried 56.1 detected errors, up 10.1% in a single year. Pages did get bigger over the same period, averaging 1,437 elements and up 22.5%, so part of that rise is simply more page to go wrong. Not all of it, though, because the errors grew faster than the pages did.
Three findings are worth naming, because behind each one sits a decision somebody made.
- Low contrast text on 83.9% of home pages, averaging 34 instances on each. Nearly all of those trace back to one choice in a palette, which is what makes contrast both the most commonly failed thing on the web and the cheapest to fix.
- A main element or main landmark on only 46.1% of them. So more than half the web offers no main landmark to jump to, which is one of the quickest routes a screen reader user has to the content, on the home page, which is where they are most likely to arrive.
- Pages using ARIA averaged 59.1 detected errors against 42 for pages without it. WebAIM is careful with that one and so are we. It does not show ARIA causing failures, because the pages using it were more complicated to begin with. What it does kill is the instinct that adding attributes is how you clean up a scanner report.
One limit belongs with every figure above, and WebAIM states it themselves. All automated tools have limits, not every conformance failure can be detected automatically, and an absence of detected errors does not mean a page is accessible. So 95.9% is a floor rather than a measurement. The true number is higher, and nobody has a way to count it.
What You Get Beyond Compliance
Three genuine side effects, none of them speculative:
- Overlap with search. Nine specific things accessibility requires are things search engines have rewarded for years: alt text, heading structure, page titles, link text, semantic HTML, captions, language attributes, responsive layout, and plain writing. Accessibility is not a ranking factor, and the work overlaps anyway.
- Fewer support contacts. Unlabelled fields and vague error messages generate confusion for everyone. Fixing them shows up as fewer "I could not work out how to" emails.
- Procurement eligibility. Public sector buyers, universities, and large enterprises increasingly ask for a VPAT before they sign. No report often means no deal, which makes accessibility a revenue question rather than a cost one.
The Arguments Against, Answered
| The objection | The answer |
|---|---|
| "We do not have disabled customers" | You cannot see the ones who left. People who cannot use a site rarely write in to say so. They use a competitor. |
| "It will make the design worse" | The constraints are contrast, target size, focus states, and reflow. Typography, imagery, motion, and personality are untouched. |
| "It is too expensive" | Retrofitting is expensive. Building it in is close to free. And in the US, small businesses can claim 50% of eligible spending back under the Disabled Access Credit. |
| "We will add a widget" | The tracker that follows these cases counted over a thousand sites sued in 2024 with a widget already running, and in 2025 the FTC ordered one vendor to pay $1 million that may be used to provide refunds to consumers and barred it from claiming its automated product makes any site WCAG-compliant without evidence. The record does not support it. |
| "We ran a scanner and it came back clean" | W3C's own automated rules say what that is worth. On 59 of its 87 live rules, a pass is recorded as meaning the criterion needs further testing, and 49 of the standard's 86 criteria have no automated rule at all. |
| "We will do it after launch" | Contrast, target size, focus states, error wording and heading levels are all settled in the design file, where fixing them costs minutes. After launch the same decisions cost a sprint and a conversation about why something signed off has to change. |
The Argument That Actually Persuades
It is not the statistics. It is not the lawsuits. It is watching somebody try to use your site and fail.
Put a screen reader user in front of your checkout and watch them stop at a button announced as "button". Nothing in a slide deck matches it. That is a large part of why our audit includes a real blind screen-reader user rather than a longer report. A recording of somebody stuck on your own checkout ends the argument faster than any number here.
One honest limit
We are not going to tell you accessibility will lift your revenue by a specific percentage, because we cannot know that and neither can anyone quoting you a figure. What we can tell you is how many people currently cannot finish a task on your site, and exactly which barrier stopped them. That is a number worth having before you argue about any other one.