Services · Focused audits
A Live Region Audit Checks What Your App Says Out Loud
For the dynamic site or app that updates without reloading, and wants to know which of those updates a screen reader user actually hears.
What We Keep Finding
Modern interfaces talk with their hands. A toast here, a counter there, three results filtered in, one item added to the cart. Sighted users see the flutter. Screen reader users hear exactly what the code announces, which by default is nothing. "Added to cart" that was never spoken did not happen.
Most teams know about politeness and stop there, and politeness is only half of it. Politeness decides when a message gets read, so whether it interrupts or waits its turn. Atomicity decides what gets read, meaning the whole region or only the bit that changed. Get that second one wrong and your cart counter announces "3" into the silence, with no word attached to say three of what. The status role carries both settings sensibly by default. A plain live region does not, and a plain live region is what most people reach for.
The second thing that goes wrong is a false negative, and it is why teams tell us the announcement is there when it is not. A live region generally has to exist in the page before the message lands in it, so a component that creates its own region and fills it in the same breath announces nothing at all. The one role that is handled differently is the alert role, which is announced even when injected. That is the difference between a toast that works and a toast that is silently perfect in the code review.
Then there is announcing twice, which is the commonest defect in the whole category and the one that looks most like diligence. Moving focus is already a change of context, which means the message was never a status message and never needed a live region on top. A form that jumps focus to the error summary and also fires a polite announcement says the same thing twice, and the second interrupts the first. It is not belt and braces. It is a bug.
One boundary worth knowing before you inventory anything, because it decides what belongs in a finding. A list of search results is not a status message. The little line saying "18 results" or "no results" is, and only if the update does not move focus or reload the page. So a filter that repaints a grid needs a count announced, not the grid.
What We Check
- Inventory every dynamic update on your key journeys, then split the counts from the content
- Test which updates announce, which stay silent, and which announce twice
- Check whether each live region existed before its message arrived
- Tune politeness and atomicity as two separate settings, since they answer different questions
- Check form feedback for errors written into the page as well as errors spoken aloud
- Verify loading states announce their start and their finish rather than spinning silently
What You Get
Every service on this site runs the same three-pass engine: an automated scan, an expert review of all 55 WCAG 2.2 A and AA rules, and a hands-on session with a professional blind screen-reader user. You get one report with every finding screenshotted, ranked by user impact, and linked to its fix. Your team fixes, we verify: the re-audit is half price within 3 months.
The format is not a mystery either. Read the sample report before you spend anything.
The Honest Limit
One honest limit, and it is the standard's position as much as ours. Announcement behavior varies across screen reader and browser pairs. The published ARIA patterns say in terms that they deliberately do not work around gaps in that support, and that you should test on the combinations your own audience runs. So we test the mainstream pairs, name their versions in the report, and record the differences we saw rather than averaging them into one answer that is true nowhere. Nothing automated helps here either, since the criterion this page is built on has no approved automated rule anywhere in the world.
What It Costs
Rapid Audit: $499, up to 10 pages you pick, report in 5 business days. Standard Audit: $1,499, up to 25 pages in 10 business days. Flat rates, no discovery calls, and a real blind screen-reader user on every engagement. Pick your pages. We bring the humans.
Worth Reading Next
Related on this site
Guides, checklists, tools, and terms that go with this service.
More in Focused audits
- Focus managementA Focus Management Audit Finds Where the Keyboard Really Is
- Touch targets & pointersA Touch Target Audit Finds the Small Buttons and Real Misses
- Timing & motionA Timing and Motion Audit Steadies an Interface That Will Not Hold Still
- Semantic structureA Semantic Structure Audit Checks the Skeleton Under the Skin