Good automation targets
- Status codes
- headers
- links
- metadata
- structured data syntax
- known configuration patterns
- repeatable performance collection
- detectable accessibility rules
Human review targets
- Business intent
- manual accessibility
- usability
- content usefulness
- security exploitability
- legal conclusions
- risk acceptance
- production change approval
Hybrid workflow
Use automation to find and organize evidence, then route the result to the appropriate reviewer. Record the reviewer, decision, and verification rather than replacing judgment with a confidence-sounding summary.
Evidence standard
- State the authorization, scope, exclusions, environment, timing, and methodology version for automated checks and human review.
- Clear accountability for the next action and the conditions that trigger escalation.
- Distinguish direct observation, calculated result, heuristic, inference, and manual judgment.
- Preserve failed, incomplete, blocked, and uncertain checks instead of silently dropping them.
Interpretation and decision record
A methodology should help reviewers reach the same bounded conclusion from the same evidence. It should not imply certainty beyond the assessed scope. Record the finding, limitation, owner, decision, remediation, and the verification required to keep website work visible, accountable, repeatable, and connected to business responsibility.
Decision and verification record
Scope
Name the website, environment, URLs, entities, templates, or user journeys included in the automated checks and human review decision.
Decision
Record the chosen action, owner, priority, dependencies, approval, and the evidence that justified it.
Verification
Confirm the owner, evidence, decision, implementation record, retest, and follow-up date are all present.
Turn automated checks and human review into an accountable record.
A reproducible automated checks and human review assessment record with scope, evidence, limitations, and verification rules.