How to Prove Accessibility Remediation
Fixing an issue and proving you fixed it are not the same thing.
Proving accessibility remediation means producing documented evidence that a specific accessibility issue was identified, assigned to a responsible person, fixed, and independently verified. It requires a chain of accountability; not just a before-and-after scan, but a record of who did what, when, and under whose authority.
The technical fix is the work. The proof is the documentation that the work happened.
What Proof Requires
- A record of the original finding; what was wrong, where, and when it was detected
- Documentation of who acknowledged the issue and who was assigned to fix it
- Evidence of remediation; what was changed, by whom, and when
- Verification by someone other than the person who made the fix
- A timestamp at every step, creating an unbroken timeline
What Does Not Count as Proof
A follow-up scan showing the issue is gone does not prove remediation by itself. It proves the issue is no longer detected. It does not show who fixed it, when, or what they changed. It does not show that anyone reviewed the fix.
Similarly, a Git commit is evidence that code changed. It is not evidence that the change was made in response to a specific accessibility finding, or that the result was verified against accessibility criteria.
Proof requires connecting the finding to the fix to the verification. Without that connection, each piece is an isolated artifact.
Why It Matters
In legal and regulatory contexts, claiming "we fixed it" is not sufficient. The question is whether you can demonstrate; with documentation; that a specific issue was addressed through a deliberate, accountable process.
This is especially important when the same issue category appears across multiple scans. A pattern of recurring findings with no corresponding record of remediation can raise questions about follow-through. Keeping the record at the time of remediation avoids this problem.
Common Misconceptions
"A clean re-scan proves we fixed it." A re-scan proves the scanner no longer detects the issue. The issue may have been fixed, or it may have moved, or the scanner's rules may have changed. A re-scan is supporting evidence, not standalone proof.
"The developer who fixed it can testify." Recollection about past work is harder to rely on than documentation created at the time. Memories fade. Developers leave organizations. A defensible record does not depend on any individual's recall.
"We track remediation in our project management tool." Project management tools track tasks. They do not enforce the separation of duties between evidence submission and review, and they do not preserve an attributed history of who checked what. Tickets get edited, reassigned, deleted. That is task management, not a record of verification.
What This Means in Practice
Remediation proof should be generated as a byproduct of the remediation process, not assembled after the fact. When a developer fixes an accessibility issue, they should submit evidence; a screenshot, a re-test result, a description of the change; through a system that timestamps it, attributes it, and routes it to an independent reviewer.
This is the difference between testing and governance. Testing finds the problem. Governance produces the documented proof that the problem was solved.
SiteRecord captures remediation records at the point of action: attributed, timestamped, reviewed, and kept in an append-only history.