Quick answer: use AI for the thinking and the writing, and use real podcast data for the facts. AI is good at turning your expertise into episode ideas and drafting pitches. It is bad at knowing which shows are still publishing, which ones interview guests, how hard each is to book, and who to email, so it invents them. Get those facts from a podcast database, give them to the AI, and approve every pitch before it goes out.
Podseeker does both in one place. It finds active podcasts that interview guests, shows how hard each one is to book and who it booked recently, finds the host's or producer's email, and drafts each pitch with AI from the show's real recent episodes and guests. Each draft comes with a fit rating and a written reason, so you know before you send whether you're a strong fit or a stretch. You approve the pitch, send it from your own Gmail or Outlook, and Podseeker follows up and tracks the reply.
This guide is for experts, founders, authors, and coaches pitching themselves. Booking clients for an agency? See how to automate podcast guest booking for clients.
What AI Does Well, and What It Gets Wrong
| Job | Let AI do it? | Why |
|---|---|---|
| Turn your expertise into episode ideas | Yes | Brainstorming angles from your bio and stories is what AI does best |
| Find podcasts to pitch | No, not on its own | AI lists shows that ended years ago, shows that never interview guests, and sometimes shows that don't exist |
| Judge how hard a show is to book | Only with data | It needs the show's recent guests and their credentials to compare you against |
| Find the host's email | Never | AI guesses addresses that look right and bounce, or go to a stranger |
| Reference a recent episode | Only one you give it | AI invents episode titles and quotes, and hosts notice immediately |
| Draft the pitch | Yes, from checked facts | It writes a good first draft when every fact in it is real |
| Draft the follow-up | Yes | One short, polite reminder is easy to get right |
| Decide what to send | No | You are the one the host will interview |
The pattern: AI writes, data decides. Most of the failures people blame on "AI pitching" are pitches built on facts the AI made up.
Step 1: Turn Your Expertise Into Episode Ideas
Hosts book episodes, not people. Before you look for a single show, work out three to five specific episodes you could deliver. "Three pricing mistakes that keep service businesses under $10,000 a month" gets booked; "I'd love to talk about entrepreneurship" does not.
Copy this prompt into ChatGPT or Claude:
Act as my podcast booking strategist. My expertise: [what you know]. My proof: [results, roles, numbers you can back up]. My stories: [two or three real experiences]. The audience I want to reach: [who]. Suggest five specific podcast episode ideas a host would want for that audience. For each, give a title, three takeaways a listener would get, and the proof from my background that makes me credible on it. Use only the facts I gave you. If an idea needs proof I don't have, say so.
Keep the two or three ideas you can talk about for 45 minutes without notes. Those become your pitch angles.
Step 2: Build a Shortlist From Real Podcasts
This is the step to keep away from AI-only research. A list of 30 shows from a chatbot looks complete, but some will be inactive, some won't take guests, and the contact details may be guesses.
With Podseeker: search your subjects and turn on the Active, Has guest, and Has email filters. Set booking difficulty to very low through medium if you're new to guesting. Or create a profile for yourself, with your expertise, credentials, and speaking topics, and Podseeker recommends podcasts matched to you, each with a fit rating. Every show comes from Podseeker's podcast database, not a marketplace that hosts have to join.
Free route: search Apple Podcasts or Spotify for your subjects, then open each show and check three things by hand: an episode in the last 60 days, recent episodes with guests, and a contact route on the show's website.
Aim for 30 to 50 candidates, then cut to your best 10 to 15.
Step 3: Check Your Booking Odds
The fastest way to waste a month is to pitch only famous shows. Booking difficulty predicts results more than most guests expect. In Podseeker's State of Podcast Pitching 2026 report (6,546 pitches sent April 11 to August 11, 2026, outcomes observed at least 60 days later):
| Booking difficulty | Pitches | Booked |
|---|---|---|
| Low | 1,758 | 5.3% |
| Medium | 3,350 | 2.4% |
| High | 1,029 | 1.2% |
| Very high | 146 | 0% |
On each podcast page, read the recent guests and their credentials. If you wouldn't look at home next to them, move the show down your list. You can also give AI the evidence and let it compare:
Here are the last five guests on [podcast] and their credentials: [list]. Here is my bio: [bio]. Compared with these guests, am I a natural fit, a stretch, or not yet credible for this show? Explain in two sentences, and name what would make my pitch stronger.
Step 4: Draft Each Pitch With AI, From Checked Facts
A good pitch is short: why this show, one specific episode idea, why you, and an easy yes. AI drafts that well when it has the right inputs.
With Podseeker: add the shows to a campaign and Podseeker drafts each pitch with AI from the show's recent episodes, recent guests, guest requirements, and host details, plus your profile and template. Each draft shows a fit rating and a one-sentence reason that compares you with the show's typical guests and booking difficulty, for example that your expertise matches the audience but the show usually books founders with named companies. Edit the draft, or skip the show.
With ChatGPT or Claude: paste in facts you have checked yourself.
Draft a short podcast guest pitch, under 150 words. Explain why I fit this specific show and propose one episode idea. Use only the facts below. Do not invent episode references, audience details, results, or familiarity with the show. If the evidence for fit is weak, say so instead of writing the pitch. Me: [bio, proof, angles]. Podcast: [description, host name, one recent episode or guest and its topic].
Before you send, check every name, episode, and claim in the draft. The prompt asks the AI not to invent anything; checking is how you know it didn't.
Podcast guest pitch template
Subject: Guest idea for [Podcast name]: [specific topic]
Hi [Host name],
Your episode with [recent guest] on [topic] stuck with me, especially [one specific point].
I'd like to suggest an episode for your listeners: [episode title]. I'd cover [takeaway 1], [takeaway 2], and [takeaway 3].
I'm [name], [one line of credibility tied to this topic].
Would this work for an upcoming episode? Happy to send a few more angles if this one isn't right.
[Name] · [website or one-sheet link]
No previous interviews? Lead with a real result, a lesson you learned firsthand, or a method you use with clients. Never invent results or credentials. For more examples, see podcast pitch examples.
Step 5: Send, Follow Up Once, and Stop When They Reply
Send from your own inbox, a few genuinely personal pitches each business day rather than a blast. Follow any guest form or submission instructions the show publishes. If there's no reply after about a week, send one short follow-up:
Write a two-sentence follow-up to my pitch below. Restate the episode idea in one line and make it easy to say no. Do not apologize or add pressure. Pitch: [paste].
With Podseeker: send from your connected Gmail or Outlook, schedule follow-ups, and Podseeker stops them as soon as the host replies.
Step 6: Track Replies and Improve
Measure pitches sent, replies, conversations, and confirmed bookings, not emails sent. After 20 to 30 pitches, ask which episode ideas and which kinds of shows got replies, and change the rest.
If you're running this on a spreadsheet, use these columns: podcast, host, contact, date pitched, angle, follow-up date, reply, outcome. In Podseeker, replies and bookings are tracked automatically, and Recommendations surface very high-fit podcasts: pitches to very high-fit podcasts booked at 6.2%, compared with 3.3% overall (methodology). Results vary by guest, show, and pitch.
To get AI's help on what to change, export your list and outcomes and ask:
Here are my last 30 podcast pitches with the show, the angle I used, and whether they replied or booked: [paste]. Which angles and which kinds of shows are working? What should I change in the next 30 pitches? Base your answer only on this data.
Which AI Tools Should You Use?
| Tool | Use it for | Cost |
|---|---|---|
| ChatGPT or Claude | Episode ideas, fit checks, pitch and follow-up drafts from facts you supply | Free plans available |
| Podseeker | Real shortlists of active shows that take guests, booking difficulty, recent guests, host and producer emails, AI pitch drafts grounded in each show's episodes, a written fit reason, sending, follow-ups, and reply tracking | $99 a month, 3-day free trial (pricing) |
| Google Sheets | Tracking, if you research and send by hand | Free |
| PodMatch | Matching with hosts who have joined its marketplace | Paid plans |
Start with a chatbot and a spreadsheet if you're pitching a handful of shows. Add Podseeker when finding active, guest-friendly shows and the right contact becomes the slow part, or when you want the drafting, sending, and tracking in one place.
A Real Example: Ben Albert's AI Assistant
Podcast host Ben Albert pitches himself with an AI assistant, Bert, that tells every host it's an AI, asks one question, and hands the conversation to Ben the moment they reply. He booked 18 podcasts in 30 days (September 2 – October 1, 2026).
"I refuse to be dishonest. He doesn't pretend to be me or pretend I listened to something I didn't. Once someone replies, I take over personally."
Read how Ben Albert uses AI to get booked on podcasts, including what hosts said and what didn't work.
Your First 30 Days
| Week | Do this | Done when |
|---|---|---|
| 1 | Write your bio, proof, and three episode ideas with the Step 1 prompt; make a simple guest page or one-sheet | Three angles you can talk about for 45 minutes |
| 2 | Build a list of 30 to 50 real shows, check booking odds, and cut to your best 10 to 15 | A shortlist with a reason for each show |
| 3 | Draft and send three to five personal pitches each business day | Every pitch checked and sent from your inbox |
| 4 | Follow up once on pitches with no reply after a week; review what got replies | A clear next angle and next 30 shows |
Treat the first month as learning what resonates. Bookings often arrive weeks after the pitch.
Using an AI Agent Instead
Comfortable with AI agents? Point Claude or ChatGPT at the Podseeker skill file with a Podseeker API key, and the agent can search for active shows that take guests, check booking difficulty and recent guests, pull contacts, and draft pitches from one request. Approve the shortlist and every pitch before anything is sent. The API is $99 a month for 2,000 credits, with 100 free credits to try it.
Frequently Asked Questions
Can ChatGPT find podcasts for me to guest on?
It can suggest shows, but it can't reliably tell you which are still publishing, which interview guests, or who to email, and it sometimes lists shows that ended or don't exist. Use it for ideas, then check each show yourself or use a podcast database such as Podseeker that filters for active shows with guests and provides host and producer emails.
Is it okay to use AI to write podcast pitches?
Yes, if every fact in the pitch is true and you would stand behind every sentence. Hosts reject pitches that are generic or reference episodes the sender clearly didn't hear, not pitches that had help with the wording.
Will podcast hosts know I used AI?
They notice generic praise, invented episode references, and the same pitch sent to every show. A draft built from the show's real recent episodes, with one specific idea and a line you added yourself, reads like a normal pitch.
Can AI book podcast interviews for me automatically?
AI can do most of the research and drafting, and tools like Podseeker can send and follow up for you, but the host still decides. Keep a person approving each pitch: one careless automated email can cost you a show you would have fit.
How many podcast pitches should I send?
Enough to learn from, at a pace where each one is personal: three to five a business day is a sustainable start. In Podseeker's data, 3.3% of pitches booked overall and 6.2% of pitches to very high-fit podcasts, so plan on dozens of pitches for your first few bookings.
What is the best AI tool to get booked on podcasts?
Use a chatbot such as ChatGPT or Claude for your angles and wording, and a podcast database built for guest booking for the facts. Podseeker combines both: real shows that take guests, booking difficulty, host and producer emails, AI pitch drafts from each show's recent episodes, and sending with follow-ups, for $99 a month.
Related Guides
- How to be a guest on a podcast: the complete guide, from finding shows to preparing for the interview.
- State of Podcast Pitching 2026: what 6,546 pitches show about booking odds.
- How to automate podcast guest booking for clients: the agency version of this workflow.
Try us risk free with a FREE 3 days trial.
