Task-Based Screen Reader Testing
Scope an actual task-based screen reader session, with recorded tester, browser, assistive technology, date, findings and retests. Human validation is separately scoped.
For Florida business websites, an automated rule result or keyboard-only pass is not a screen reader session. We can scope a human session around agreed customer tasks, record the environment and document what was actually observed.
Agree on tasks and the test environment
Before a session, agree on pages, tasks, access, safe test data and vendor handoffs. The delivered record identifies the tester, actual session date, browser, operating environment and assistive technology used. NVDA or VoiceOver may be appropriate examples depending on the agreed audience and environment; neither is implied to have been used unless the delivered record says so.
Automated axe/DOM checks can flag code patterns, and keyboard-only checks can expose focus problems. Neither is a substitute for a person operating the task with a screen reader. See our testing method.
Listen to the task, not just the page
Agreed sessions can follow navigation menus, link context and image alternatives into a form, then check labels, instructions, error recovery and confirmation. For a dialog, assess announcement, focus order and return; for live updates, check whether status is conveyed without noise. Record where a vendor widget or hosted checkout takes over and whether the agreed task could continue, without claiming to certify vendor software.
Choose a representative journey rather than treating a clean homepage as proof about every template. Custom authenticated or payment journeys need separate authorization and scope.
Findings, repair and retest
The deliverable distinguishes an observed barrier from a recommendation, with task steps, environment, date, expected behavior, evidence and ownership or vendor limitation. If a repair is agreed, retest the relevant task and record the outcome and remaining limitations; a finding is not resolved merely because a later scan did not detect it. A separately scoped independent human validation can review repaired tasks.
Read what a remediation record includes and how WordPress source repairs are approached.
Request the right scope
Human screen-reader testing is separately scoped, not included by default in the starting website remediation price of $2,449 for up to 15 pages after confirmation. Monitoring starts at $49/month for agreed automated checks, not human sessions or repairs. Request a free public-page scan for initial signals or request a human-testing estimate. See the service overview.
Sources
Request a free accessibility scan
Selected public pages only; not human review or recurring enrollment. Request a free scan.
Request a scoped estimate
Human screen-reader testing is separately scoped. Tell us the tasks, access constraints and assistive-technology audience to plan a session.
Frequently Asked Questions
Is an axe scan or keyboard check a screen reader test?
No. Both can identify important barriers but neither means a person completed an actual task using assistive technology. Agree on a human session and record its tester, date, browser and assistive technology.
Which screen readers do you test with?
NVDA and VoiceOver are examples that may be chosen according to the agreed audience, browser and environment. The delivered record states what was actually used; this page does not claim a session has already taken place.
Will you test a third-party checkout?
Only if the specific journey, authorization and safe test data are agreed. We can document the handoff and an observed barrier but cannot claim to control or certify vendor-owned code.
Is human validation included in the starting remediation price?
No. Human screen-reader sessions and independent validation are separately scoped. Website remediation starts at $2,449 for up to 15 pages after scope confirmation; monitoring starts at $49/month for agreed automated checks.