# Research how leaders demo accessibility

> Pull a11y walkthroughs across products and study the language and flows — spec your audit against the best.

- For: PM
- Category: Product & UX Research
- MCP tools: deep_search, library_import_youtube_video, get_youtube_transcript
- HTML version: https://veridive.com/mcp/use-cases/a11y-demo-research

## The problem

Accessibility work usually starts from compliance documents — WCAG checklists, audit templates — which tell you what to satisfy but not what good looks like *in a product like yours*. The gap between "technically compliant" and "actually usable with a screen reader" is where most a11y efforts land short, and no checklist closes it.

What closes it is studying the leaders: companies that treat accessibility as craft demo it — walkthroughs of their screen-reader flows, conference talks on their a11y engineering, disabled users reviewing what actually works. That corpus shows the state of the art as *practiced*: the flows, the patterns, the language teams use when they take this seriously. Specs written against it aim at good, not just at compliant.

## How veridive MCP helps

**deep_search** collects the practice corpus: a11y walkthroughs and engineering talks from product teams, plus — the layer most teams never see — reviews by disabled users evaluating real products with real assistive tech. Transcript search finds the demonstrations, not just the mission statements.

**library_import_youtube_video** and **get_youtube_transcript** make it studyable: import the strongest sessions, pull the transcripts, and extract the working material — how leaders structure screen-reader flows, the exact language of their announcements and labels (transcripts capture what the screen reader actually said), the patterns disabled reviewers praise versus route around. Your spec inherits practiced patterns; your audit gets user-validated criteria beyond the checklist.

## How to run it

1. **Connect veridive MCP** — One-time setup: add the veridive MCP server to Claude, ChatGPT, Cursor or any MCP client and generate your API key. Takes about two minutes.
2. **Collect the practice corpus** — Leader walkthroughs, a11y engineering talks, and assistive-tech user reviews — the craft layer.

   ```
   Find accessibility walkthroughs and engineering talks from product teams known for a11y, plus reviews by screen-reader users of <product type> — use veridive's deep_search.
   ```
3. **Extract the working patterns** — Import and transcript the best sessions — flows, announcement language, praised patterns.

   ```
   Import these and pull the transcripts — extract how they structure screen-reader flows and the exact announcement/label language used: <video URLs>
   ```
4. **Spec against the best** — Your flow, specified with the studied patterns — beyond the checklist, toward the craft.

   ```
   For our <flow>, draft a11y spec guidance based on these patterns: flow structure, announcement language, and what assistive-tech users praised or avoided.
   ```

## Problems it solves

- Checklists define compliant; nothing internal defines good — and the gap is where users struggle.
- The craft layer (leader demos, a11y engineering talks) exists and never reaches spec-writers.
- Disabled users' product reviews — the ground truth — are absent from most teams' research diet.
- Announcement and label language gets invented per-team when practiced patterns are documented on tape.

## Why it's valuable

- Specs aim at practiced excellence, not checklist minimums.
- Transcripts capture the exact announcement language — the hardest detail to get right blind.
- Assistive-tech user reviews inject ground truth the compliance audit can't.
- The corpus doubles as training material for engineers new to a11y work.

## Example prompts

```
A11y walkthroughs from strong teams + screen-reader user reviews of <product type> — use veridive's deep_search.
```

```
Import these; extract flow structures and announcement language: <URLs>
```

```
Draft a11y spec guidance for our <flow> from these patterns.
```

## FAQ

**Does this replace an accessibility audit?**

No — it upgrades what the audit measures against. Compliance review stays; this supplies the craft criteria and user-validated patterns that turn a passing audit into a usable product.

**Why do the transcripts matter so much here?**

Because a11y quality lives in spoken detail: what the screen reader announces, in what order, with what labels. Transcripts of walkthroughs capture that verbatim — the one artifact screenshots can't provide.

**Where do disabled users' reviews fit in?**

They're the ground truth: patterns leaders demo and users route around are failed patterns. Weight the review layer highest when it conflicts with the demo layer.

## Related use cases

- [Survey design patterns across product demos](https://veridive.com/mcp/use-cases/design-pattern-survey.md) — Collect demo videos across your space and query interaction patterns — a design review across the whole market.
- [Treat tutorials as free usability tests](https://veridive.com/mcp/use-cases/tutorials-as-usability-tests.md) — Every workaround a tutorial teaches is a usability bug report; collect and query them across your product's ecosystem.
- [Mine review videos for user pain](https://veridive.com/mcp/use-cases/user-pain-mining.md) — Reviewers narrate friction in real time; harvest the "why is this so hard" moments across your category.

---

Part of the veridive MCP use-case library: https://veridive.com/mcp/use-cases.md
Server docs for agents: https://veridive.com/mcp.md