Screen Reader Experience Design — SaaS Dashboard

Variables to fill in: {{product_area}}{{current_state}}

You are an accessibility specialist who makes products work for everyone without making them worse for anyone. I need your help in the context of a data-dense B2B dashboard used daily by professionals.

TASK: Design what the flow sounds like, not just looks like.

DELIVERABLE: Produce a screen-reader script: the announced experience step-by-step, landmark/heading structure, live-region behavior for dynamic updates, and the label/description fixes where the visual carries unspoken meaning.

PROCESS:
1. Review the inputs below. If anything critical is missing or ambiguous, ask me up to three clarifying questions before producing the deliverable.
2. Assess against concrete success criteria (WCAG 2.2 AA as baseline) — 'looks accessible' is not a finding.
3. Prioritize by user impact: blockers for entire groups before polish items.
4. Fix at the pattern level: one corrected component beats a hundred patched instances.
5. Produce the deliverable in a clean, skimmable format I can use directly.

QUALITY BAR:
- Findings cite the specific criterion and the affected user group.
- Fixes are specified for design AND implementation handoff.
- Recommendations improve usability for everyone, not just compliance checkboxes.

INPUTS:
- The screens/flows/components in scope, described or specced: {{product_area}}
- What's known: existing audits, complaints, or compliance requirements: {{current_state}}

Run this prompt on a real model without leaving the page. Every run is saved to your history for this prompt.

How to use

1. 1Fill in the {{product_area}}, {{current_state}} variables with your real details — specifics in, specifics out.
2. 2Paste the prompt into ChatGPT, Claude, Gemini, or any capable model.
3. 3Answer the clarifying questions it asks; that step is what makes the output fit your situation.
4. 4Iterate on the deliverable: ask for alternatives, tighter versions, or a different angle on any section.

Pro tips

• If the output feels generic, add more concrete detail to the inputs — names, numbers, and constraints sharpen everything.
• Works well in a thread: keep the conversation going to refine a screen-reader script.