Handling accessibility demand letters
This must be the “demand letter” season. Maybe because all the lawyers are working from home due to COVID-19 with courthouses being closed? I’ve had multiple people reach out to me this week asking me how to respond. Remember this is NOT legal advice. But this is what I do.
One: Do NOT think that installing an overlay/plugin/tool is going to magically solve your accessibility issues
TL; DR (though you really need to read the details)
If it were true that an overlay would eliminate your risk, why wouldn’t there be overlays on EVERY site?
You can’t automatically fix what can’t be automatically detected. Even the most optimistic estimate of how much can be detected is 50 %. My estimate, and the estimate of the colleagues that I respect, is 30–40 %.
Overlays treat the symptoms, not the root cause.
Overlays never “solve” the problem the way the USERS want it solved.
Overlays only “work” in the web world, when the world is moving to mobile.
Overlays may impact your site’s security and performance. At a minimum, your IT department is going to have to constantly monitor for this
You are going to be paying for that overlay service forever until you actually SOLVE your accessibility problem!
A LinkedIn connection just reported to me the results of some superficial reviews they performed of the websites of every client that a major overlay company proudly claims as customers.
Not a single one of them passed even the WAVE tests, much less manual review. And the WAVE tests are what the plaintiff’s lawyers largely use. How do I know that? 1) They are free and lawyers don’t like to pay for things, and 2) WAVE results are in the pleadings of most cookie-cutter lawsuits.
These are not false positives, these are real and significant issues.
In the end, the use of an overlay will NOT be sufficient to make your site compliant and therefore will NOT help you avoid lawsuits and legal judgments.
Two: Do not fall into the trap of thinking insurance or indemnity is going to cover your costs/losses
A second LinkedIn connection just forwarded me a link to a company that is offering free risk assessments and $50,000 in litigation insurance. Let’s break that down. How far is that $50,000 going to carry you? It will get you to the airport (maybe) in your around-the-world accessibility journey. Settlement costs are typically about 5 % of the overall cost of litigating an inaccessible website.
MIT just shelled out $1,000,000 for the plaintiff’s fees alone in their four-year litigation journey involving the lack of closed captioning and reliance on auto-captions on MOOC sites. This means they probably spent at least that on their own legal fees. That $2,000,000 could have paid for almost 27,000 HOURS of the captioning they were trying to fight having to provide. Assume arguendo that co-defendant Harvard’s fees were equal. I have zero reason to believe they weren’t equal because they were co-defendants in the same case. Harvard actually held out for months longer than MIT did, so their legal fees were probably higher. That’s more than 50,000 hours of captioning they could have paid for. Instead, MIT and Harvard effectively flushed $4,000,000 down the toilet.
Most contracts do not have clauses requiring service providers to deliver WCAG-compliant goods and services. When there is not a specific clause instead you will be forced to rely on vague fall-back language about “goods and services having to abide by federal, state and local laws.”
Here’s the problem with that vague fall-back language — The DOJ never officially adopted WCAG as a standard. WCAG as a standard is coming strongly out of equal access arguments in case law. I wouldn’t want to be the one claiming that case law is included in the vague fall-back contract language I cited above, so good luck recouping your costs from an indemnification claim if that is your situation.
You may have slightly better luck on indemnification if you are based in California. The California government has officially integrated WCAG 2.1 Level AA into its regulation as its accessibility standard in some areas such as public-facing websites and CCPA.
Three: Do NOT get a “free risk assessment”
Your free risk assessment IS NOW DISCOVERABLE by the plaintiff’s legal counsel if you didn’t get it as attorney work product and didn’t get it under a non-disclosure agreement.
People are always attracted to getting something for free, DON’T DO IT.
Largely the “free” risk assessments are thinly guised attempts by accessibility consultancies to sell remediation services and overlays. The free risk assessment:
likely doesn’t screen out false positives
likely didn’t include any of the issues that can only be uncovered by intensive manual testing (spoiler: this is usually 2/3 of the total bug count)
likely doesn’t include your highest risk areas which are communications, complaints, and internet transactions — none of those can be tested in an automated manner.
And while not doing the things you really need, a free risk assessment is going to leave a discoverable electronic paper trail that the plaintiff’s attorneys will gobble up and use against you.
Two takeaways here:
Free risk assessments will not provide you a complete picture of your actual risk in any way shape or form, no matter how the providers advertise them.
Free risk assessments can do you VASTLY more harm than good if they aren’t done correctly.
Four: Do not feed the trolls
Responding to a demand letter without knowing your actual situation is literally the worst thing you can do.
Consider it a “shot across the bow” and act accordingly.
Protect as much as you can under Attorney Work Product
Under Hooter’s case, you can get multiple demand letters from multiple serial plaintiffs and have to pay settlements out to each and every one. Unless class action litigation is filed and settled, accessibility defense is literally death by 1000 papercuts. Don’t hand the trolls a ream of paper. Silence is golden.
But it’s not all doom and gloom. When you get one of these letters (notice I said when, not if?) there are some things you can do.
One: Figure out your true accessibility status
Do you have an accessibility policy?
What does it say?
How are you measuring whether or not you are doing what your policy says you are going to do
Do you have someone in charge of accessibility who actually has the chops to know what they are doing? Don’t make the mistake GNC did. You can’t wave a magic wand over a random IT person, send them to a couple of webinars, and claim they are an expert.
How do you handle accessibility outside of IT — for your true disability/accessibility story you need to check HR, D&I, procurement, training, customer support, and tech pubs, for starters.
Two: Build an actionable remediation roadmap
The key word here is “actionable”
It has to be resourced. You can start making things accessible on a low budget, but you can’t do it on a zero budget.
It has to be prioritized. An essential component is messaging from the top that accessibility is now important despite previous messaging. Some flavor of: “We did this before. We were wrong. Here is our new approach going forward” is the best way to do that.
It has to be driven. Put someone who actually cares about accessibility for a deeply personal reason in charge. They are the kind of people that will make it happen despite all the setbacks and blatantly discriminatory behavior they are going to see along their remediation journey
It requires training. Make sure everyone who touches the code, designs, documentation, and services knows what they need to know.
Three: Call out your desired standard in every contract you execute in the future
That will help you with your indemnification arguments that I called out above if it turns out a vendor is responsible for the inaccessible code the demand letter is complaining about. Warn current vendors that this clause will be coming on renewal and if they don’t agree, another vendor who WILL agree will be located.
Four: Add Accessibility to all your OKRs and Customer Journeys
Employees do what they are rewarded to do. If you reward them for launching features, they will launch features. But those launched features might not be accessible. If you reward them for creating accessible software, they are more likely to dedicate themselves to learning about accessible software and then execute using that new knowledge.