ChatGPT Prompts for Podcast Hosts
Guest research dossiers, episode outlines that don't sound scripted, episode titles, audiogram pull-quotes — built for the recurring prep and promotion work around every episode.
Podcast hosting involves more recurring research and writing work than most people realize when they start. Around every interview sits prep (guest research, outline, intro draft) and post-production writing (show notes, episode title, social promotion). These prompts are built to speed up that prep + promotion workflow while preserving the unique voice of your show. The interview itself stays where it belongs: in your hands. The sample outputs below use a fictional guest so you can see the shape of each output without anyone's real record being invented.
Key Takeaways
- Most of the work in an interview episode happens before and after the recording. Prompts should target the prep + promotion sandwich (guest research, outline, intro, show notes, social clips), NOT the interview conversation itself. The interview stays yours.
- Guest-prep prompts should mine the guest's LAST 3-5 interviews FIRST, then ask 'what questions has this guest been asked to exhaustion, and what would they genuinely enjoy talking about that no one asks'. That single framing produces better questions than generic 'guest interview prep'.
- Show-notes prompts should be structured: timestamp navigation, the guest's full name, 3-5 key takeaways, resource links. Show notes are where listeners go for links and timestamps, and Apple accepts up to 10,000 characters in an episode description — use the room for substance, not filler.
- Social clips: prompt for TOP QUOTABLES from the transcript (paste it in), then generate caption + hook + hashtag set per clip. Several clips spread over the following weeks give an episode several chances to be found, where a single launch-day clip gives it one.
“Also, don’t repeat the title of your show within your episode title. Separating episode and season number from the title makes it possible for Apple to easily index and order content from all shows.”
Guest Research Dossier (Public Materials Only)
Build a guest research dossier I can use for episode prep. Focus on what's PUBLICLY available and verifiable. The goal is to know enough to ask sharp questions, not to know everything about them. Guest name: [name] Guest's primary public role: [what they're known for — author, founder, executive, expert, creator] Episode topic / angle: [the specific angle we're recording] Episode length: [30 / 45 / 60 / 90 min] My show's tone: [direct/journalistic / warm/conversational / debate/challenging / educational/explanatory] Source materials to check: [list any books, talks, articles, podcast appearances you know exist] Output the dossier as: **The 30-second version:** Who is this person to my audience? **Key public claims (3-5):** Specific positions or arguments they've made publicly — with the source or year **Recent activity (last 12 months):** What have they done, said, or published recently that's relevant? **Where they've contradicted themselves or evolved:** Public moments where their view changed (this is where the BEST interview questions live) **Topics they've talked about A LOT:** What they default to (so I can steer AWAY from these and toward fresher ground) **Topics they've NEVER addressed publicly:** Areas of expertise they've never been interviewed on **Personal context safe to mention:** Anything public-domain about their background that informs their work (don't speculate or include rumor) **Three uncomfortable but fair questions:** Questions a thoughtful interviewer would ask, that they probably haven't been asked before **Three "easy" questions to open with:** Conversational openers to warm up the conversation before the harder questions **Production note:** The 2-3 things I should mention I've checked / read before the recording starts so they know I did my homework
Episode Outline That Sounds Like a Conversation
Build an episode outline I can interview from. The output should look like a roadmap, not a script — I need flexibility to follow where the guest takes the conversation, while having structure to return to. Episode title (working): [title] Guest: [name and 1-sentence credential] Episode length: [30 / 45 / 60 / 90 min] Show format: [solo interview / co-host interview / panel / educational solo episode] The ONE thing the audience should remember after listening: [the takeaway in one sentence] The three things the audience should LEARN: [bulleted facts/insights] Specific question I'm hoping the guest will answer (if any): [the one question only this guest can answer] Output as: **Cold open (60-90 sec):** A specific question or moment that hooks the listener immediately. NO show intro yet. Just the hook. **Intro (30 sec):** Show intro + guest intro. Brief and energetic. **Section 1 (10-15 min): [Open ground]** - 2-3 opening questions to warm up the guest - The transition cue to move to Section 2 **Section 2 (15-20 min): [The core territory]** - 3-4 questions covering the main subject - ONE follow-up question I should be ready to ask based on a likely answer - The transition cue to move to Section 3 **Section 3 (15-20 min): [The depth or challenge]** - 2-3 sharper questions that go deeper than what's been said publicly - The ONE question I'm hoping they'll answer - The transition cue to wrap **Closing (5-10 min):** - Reflection question (gives them space to take the conversation wherever they want) - The "where to find you" practical info - A specific call-to-action for the audience **Buffer questions (3-5):** Extra questions I can pull if conversation runs short **Cut signals:** Three things that would tell me a section has gone too long and I should move on — be specific (e.g., "if the guest gives a 5+ minute answer with no concrete examples, ask one followup and move to next section")
Episode Title and Subtitle (3 Versions)
Write 3 versions of an episode title and subtitle for podcast platforms. The title appears in the listener's feed; the subtitle appears in show notes and on social. Both need to do work. Show name: [name] Guest name (if interview): [name] Episode topic in 1-2 sentences: [what the episode is actually about] The audience this episode is FOR: [be specific — e.g., "B2B SaaS founders considering AI tools", "creative writers who feel stuck", "lawyers under 35"] Show's title convention: [e.g., "GUEST NAME — TITLE" / "TITLE (GUEST)" / Punchy standalone titles] Episode genre: [interview / solo / panel / educational / news/timely] Output 3 version pairs (title + subtitle), each optimized for a different style: **Version A — Curiosity / Question lead:** - Title leads with a question or surprising claim - Subtitle elaborates **Version B — Specific outcome lead:** - Title states what listeners will get - Subtitle previews the structure **Version C — Name/Authority lead (only useful for interviews with known guests):** - Title leads with the guest's name and a sharp angle - Subtitle elaborates For each version, include: - Title (short, with the words that matter first — and no episode or season number, which belong in their own feed tags) - Subtitle (under 150 characters) - Which audience segment this version is best for - Which platforms it suits best (Apple Podcasts, Spotify, YouTube) and why
Audiogram Pull-Quote Selection
From an episode transcript, identify 5 pull-quotes worth turning into audiograms (30-90 second clips for social). Episode transcript: [paste the full transcript or section] Episode topic: [1 sentence] Audience: [target listener for this episode] Audiogram strategy: [single platform — e.g., Instagram Reels / multi-platform / clip for show notes only / promotional teaser for an upcoming episode] For each of the 5 pull-quotes, output: **Pull-quote #1:** - **Quote** (verbatim from transcript, with speaker name) - **Length** (seconds it would take to read at natural pace) - **Why this works**: what makes this clip-worthy (surprising claim / specific data / vivid metaphor / contrarian view / emotional moment) - **Best platform**: where this should run (Instagram Reels, TikTok, LinkedIn, X, YouTube Shorts) - **Pre-roll setup** (1 line of caption text that primes the listener for context) - **Post-roll call-to-action** (1 line: where to listen to the full episode) - **Visual style** (if making an audiogram with text overlay, what should the text say in 5-7 words?) After the 5 pull-quotes: **The clip NOT to use:** The most viral-seeming moment that you should actually skip, with the reason (e.g., requires too much context, takes out of context, harms guest relationship, exposes a confidential moment)
The Numbers Behind These Prompts
Every figure below cites a primary source. Click through to verify.
The most Apple Podcasts Connect accepts in an episode's Name field. It's a ceiling, not a target: the Episode Title prompt asks for short titles because listeners scan them in a feed, and Apple's RSS guide separately says not to put episode or season numbers in the title — those go in their own tags.
Source: Apple Podcasts for Creators — Create an episode2026
How much text Apple accepts in an episode's description tag, with rich text and some HTML allowed (the show-level description is capped at 4000 bytes). That's room for timestamps, takeaways and resource links: feed the show-notes structure above, the guest's full name and your transcript into ChatGPT rather than asking it for "show notes" cold.
Source: Apple Podcasts for Creators — A Podcaster’s Guide to RSS2026
Apple's recommended maximum for chapter titles in a podcast chapters file. If you turn the Episode Outline's sections into chapters after recording, ask ChatGPT for chapter names under 45 characters alongside the timestamps.
Source: Apple Podcasts for Creators — A Podcaster’s Guide to RSS2026
How to Use These Prompts
Run the Guest Research Dossier 2-3 days before recording, then the Episode Outline 24 hours before. After recording, use the Pull-Quote Selection prompt on your transcript to identify your audiogram clips, and the Episode Title prompt to draft 3 title variations for your headline test. Always verify specific quotes and statistics in the dossier output against original sources before using them on air. Prompt Anything Pro lets you trigger these prompts on any text you've selected (interview transcripts, guest research, episode notes) without copy-pasting between tabs.
Need More Prompts?
Get personalized AI suggestions for additional prompts tailored to your specific needs.
AI responses are generated independently and may vary
Frequently Asked Questions
Sources & Further Reading
- A Podcaster’s Guide to RSS — episode title, episode number and season tags, and artwork requirements — Apple Podcasts for Creators (accessed Oct 22, 2026)
- Create an episode — required tags and the character limits for episode name and description — Apple Podcasts for Creators (accessed Oct 22, 2026)
Run These Prompts Inside Any Webpage
Prompt Anything Pro lets you run these prompts on any text you've selected — interview transcripts, guest research, episode notes — without copy-pasting between tabs.