AI show notes that do not read like AI: the structure that earns episode-page rank
The 6-section show-notes template that ranks for episode keywords. Plus the AI tells to ban in podcast show notes specifically.
The direct answer
The 6-section show-notes template: hook (2-3 sentences), guest intro (2 sentences with link), key topics covered (5-7 bullet timestamps), 3-5 quotable moments, links mentioned, CTA. AI-generated show notes default to a "summary + chapter list" structure that doesn't rank — the 6-section template ranks because it matches what podcast-search-intent users actually want.
Show notes are the SEO surface of every podcast episode. Done well, they rank for "[guest name] interview", "[topic] podcast", and dozens of long-tail queries. Done badly (or with default AI output), they're a bullet list nobody reads.
The difference between default AI show notes and ranking show notes is structure. Most AI tools (Castmagic, Descript, etc.) default to a summary-plus-timestamps format. That format reads like AI to a human and ranks badly because it doesn't match query intent.
The 6-section show-notes template
Hook (2-3 sentences). The one-line claim or contrarian framing from the episode. Same shape as a newsletter opener — gets the reader to want to listen.
Guest intro (2 sentences, with link). Who the guest is, what they're known for, link to their primary online surface (LinkedIn or website). Skip if the host is solo.
Key topics covered (5-7 bullets, each with a timestamp). The "table of contents" of the episode — readers scan this to decide whether to listen and where to jump.
3-5 quotable moments (verbatim, attributed). Direct quotes from the episode, each one a self-contained idea. These are what readers screenshot and share on social — they MUST be verbatim, not AI paraphrases.
Links and resources mentioned. Every product, book, person, study referenced in the episode. This is high-SEO-value because it earns inbound links from the referenced sources and ranks for related queries.
CTA. Subscribe + leave a review + follow on social. Brief, not pushy.
Total length: 400-600 words. Reading time: 2-3 minutes. Long enough to rank, short enough to actually be read.
AI tells specific to podcast show notes
"In this episode, we discuss..." — kill this phrase. Every AI tool defaults to it. Replace with the actual claim of the episode.
"Tune in to learn..." — same kill list. This is podcast-tutorial-template language.
"Whether you're a beginner or expert..." — addresses-everyone framing kills specificity.
Generic chapter titles ("Introduction", "Main Discussion", "Conclusion"). Each chapter should be named for its actual content.
Promotional ending ("Don't forget to subscribe and leave a 5-star review!"). Most podcasts have moved on from this — it reads as performative.
What ranks vs what doesn't
Episode pages rank for two query types:
Guest-specific queries: "[guest name] interview", "[guest name] podcast", "[guest name] [your podcast name]". These rank based on URL slug, page title, and guest name density in show notes.
Topic queries: searches matching the episode's core theme. These compete against general blog content. Rank by: matching H2 structure, semantic richness (related terms), and inbound links.
The 6-section template hits both. The default AI summary-and-chapter format only hits the first.
The Persona Brief override for show notes
Generic show-notes AI doesn't know your podcast's voice. Override:
Add to the Persona Brief: "Show notes must use the [your name]/[your podcast] voice, not the generic-podcast voice. Never use phrases from the banned list (below)."
Banned list: "In this episode", "Tune in", "Whether you're", "Don't miss", "packed with", "deep dive", "covering everything from X to Y".
Required structure: "Always use the 6-section template (hook / guest / topics / quotes / links / CTA). Each quotable moment must be verbatim from the transcript, attributed to the speaker."
Tone calibration: "Match the energy of [reference podcast]'s show notes. Specific, opinionated, no fluff."
This Persona Brief addition pays off forever. Apply it once; every episode's show notes match your podcast's voice automatically.
Frequently asked questions
How long should podcast show notes be?
400-600 words. Long enough to rank for episode-specific queries, short enough to actually be read. Shorter than that, you under-cover the topic; longer, the reader scrolls past.
Should show notes include a full transcript?
Optional. Some podcasters publish full transcripts on a separate URL linked from the show notes page. Embedding the full transcript in show notes works for SEO but hurts readability.
Which AI show notes tool is best?
Castmagic for show notes specifically. Descript if you edit in Descript anyway. Kompozy if you want show notes as one of many fan-out outputs on the same Persona Brief. All three need the Persona Brief override to avoid generic AI tone.
Do quotable moments really need to be verbatim?
Yes. The whole point of pull quotes is they sound like a human said them. AI paraphrases sound like AI. Always pull quotes directly from the transcript with light cleanup (filler words removed) only.
Where should I publish show notes?
On your podcast's own website, with a unique URL per episode. Most podcast hosts (Buzzsprout, Captivate, Transistor) auto-generate an episode page; you can override the show notes content via their interface.
How do I prevent AI show notes from sounding generic?
Persona Brief override with banned phrases + required 6-section structure + tone-calibration against a reference podcast. Without this, every AI show-notes tool defaults to "Inc.com explainer" voice.
AI Brand Voice & Persona — Without a Persona Brief, every AI output averages to the LLM default voice. This is the 5-section methodology that makes 100+ AI-generated posts feel like one human author wrote them.
AI Content Repurposing — The complete methodology for turning one source into 25-35 pieces of native-format content across every platform — without producing AI slop.