Get found & cited

An audio or video file alone gives a search or answer engine very little quotable evidence. Indexing turns each episode into a server-rendered document, a structured knowledge graph, and machine-readable interfaces that can resolve a claim to the exact moment it was spoken.

This applies to both source forms equally. A YouTube channel's videos compete inside YouTube on titles and thumbnails, but the conversation itself is invisible to Google and answer engines until it exists as structured text — the surfaces below are what make a YouTube episode quotable outside YouTube, exactly as they do for an RSS feed.

PodHood creates the eligible surfaces; no product can guarantee that Google, ChatGPT, Perplexity, or another engine will crawl, rank, or cite a particular episode. Coverage grows as the catalog is indexed and external systems revisit it.

What an indexed episode publishes

For search engines

• A canonical HTML episode page with summary, Q&A, chapters, Key moments, speakers, entities, and the full Transcript.
• Schema.org JSON-LD centered on PodcastEpisode, including speakers/entities and timestamped Clip nodes for chapters and Key moments.
• FAQ-shaped structured data when the Brief contains Q&A.
• Topic/Person/Company/Product Hubs that aggregate the Channel's published evidence about a subject.
• A Channel sitemap.xml and crawlable internal links.

For answer engines and agents

• A Markdown twin for every episode: append .md or request Accept: text/markdown.
• Channel llms.txt containing the published catalog and MCP endpoint.
• An anonymous, read-only MCP search tool that returns grounded, timestamped quotations.
• A credential-gated creator REST API (with an OpenAPI description) for the channel's own tooling.
• MCP discovery at /.well-known/mcp and API discovery at /.well-known/api-catalog.
• A discoverable Agent Skill index at /.well-known/agent-skills/index.json on the PodHood platform origin.

For precise citations

Chapters, Key moments, transcript matches, MCP results, and REST search results carry URLs to the exact timestamp. That lets a citation land on the supporting moment instead of dropping a reader at the top of a two-hour recording.

Why the knowledge graph matters

A transcript makes the words crawlable. The Brief and knowledge graph make the content navigable and reusable:

• Speaker identity tells a system who made a claim.
• Entities and topics connect one episode to the rest of the archive.
• Hubs create stable subject pages.
• Q&A phrases content in the form people commonly ask an assistant.
• Timestamped Key moments provide self-contained, citable claims.

This is why the team's corrections in the Studio matter to discovery. Fix high-value speaker and entity errors before promoting a Hub or testing citations.

Pre-launch verification

Pick one representative episode and run this sequence:

1. Public HTML — open the signed-out episode page and check title, summary, chapters, speakers, Transcript, Q&A, and canonical host.
2. Timestamp — copy a Key moment or Transcript link and confirm playback/reader focus starts at the intended moment.
3. Markdown — append .md; confirm the important text and moment URLs are present without relying on the visual UI.
4. Channel index — open /llms.txt; confirm the episode and MCP endpoint appear.
5. MCP — use Connect AI, ask one real question, and open a returned citation.
6. Structured data — run the episode URL through Google's Rich Results Test and inspect the detected PodcastEpisode/Clip data.
7. Sitemap — confirm the episode or its Hub appears under the canonical host's sitemap.

Repeat this with one episode that has multiple speakers and one that contains a high-value recurring topic.

After launch

• Add the canonical Library host to Google Search Console and submit its sitemap.
• Index a coherent topical cluster, not only isolated episodes; the Hubs become more useful as evidence accumulates.
• Review MCP/Search results for the questions the team actually wants the show to answer.
• Correct misattributed speakers and duplicate entities in the Studio.
• Re-index older high-value episodes that predate newer Brief fields such as Q&A.
• Watch the Studio Subscribers roster as a separate signal of human audience growth.

Fresh pages may take time to be crawled. A citation test that fails immediately after publication is not proof that the page is broken; first verify the HTML, Markdown, llms.txt, sitemap, and MCP paths directly.