Representative audit engagement showing what a product team receives before any remediation starts: sampled coverage, evidence, severity, selectors and acceptance tests.
Keyboard, form semantics, landmarks, dynamic-status checks, automated DOM assertions, manual acceptance script
CHALLENGE
What was wrong
The product team had a release candidate but no defensible accessibility baseline. Findings from spot checks were scattered and not actionable enough for sprint planning.
Define six representative screens and two user journeys
Run keyboard and semantic structure checks
Capture 14 reproducible findings
Assign severity and exact selectors
Write one acceptance test for every finding
Produce a developer-ready audit board and PDF
VERIFIED FIXTURE RESULTS
What actually changed in the packaged evidence
CHECK
BEFORE
RESULT
Screens with defined audit coverage
0 / 6
6 / 6
Findings with exact selector
0 / 14
14 / 14
Findings with acceptance test
0 / 14
14 / 14
Major blockers isolated
Unknown
3
Release baseline
Absent
Versioned audit v1.0
14
found / baseline
Audit output - remediation not included
resolved / produced
14 prioritised findings handed over
remaining / handover
Claim boundary: results above are measurements from the representative fixture/evidence set included with this case. They are not claimed client conversion, revenue, legal-conformance or historical-performance figures.
The agency owns the client relationship, timeline and final presentation. SYNTHESIS owns the agreed specialist production and verification. Report branding and client-contact boundaries are agreed before work starts.
What stays explicit
Scope and exclusions
Dates and rates
Testing / evidence boundary
Client-contact route
Known limitations
Retest or audit status
Have a project with the same problem?
Send the staging URL, document or procurement request. We will scope the smallest useful production job and the proof you will receive.