Accessibility Prioritization Method

Prioritize repairs using observed barriers, task impact, frequency and shared causes. This framework does not predict lawsuits or settlement costs.

Updated September 11, 2026 · whatisADA / Grow Wild Agency

Start with an observed barrier

Describe the task, the affected interaction and the consequence for a user. A keyboard trap in checkout and an unclear label on an infrequently used reference page require different repair priorities. Record the actual evidence instead of inferring accessibility from an industry, revenue figure or traffic estimate.

Use a transparent decision record

Identify the task consequence: blocked completion, substantial difficulty or a smaller loss of clarity.
Record how often the barrier appears and whether it affects a shared component.
Identify a practical workaround, if one exists, without assuming every visitor can use it.
Assign a repair owner and acceptance condition. Track a temporary alternative separately from the permanent repair.
Retest the original scenario and related component states before closing the issue.

Example: choose the first repair

Suppose a modal traps keyboard focus on every product page while an older article has a vague link label. Fixing the modal removes a repeated task blocker. The link still needs repair, but the decision record explains why it follows. If that link is the only route to emergency information, the task impact changes and the priority should be reconsidered.

What the framework cannot establish

It does not assign a probability of litigation, forecast a settlement or replace legal advice. Filing totals describe reported cases in their source population; they do not provide a calibrated probability for an individual site. WCAG conformance levels also are not a universal severity scale. Keep legal applicability, technical findings and operational repair priority as separate documented decisions.

The former point-based risk presentation has been replaced by this planning method and the Accessibility Readiness Planner. No validated predictive dataset is published for that former score.