Where to Find Great Podcast Guests You Haven't Met
To find podcast guests you haven't met, work five sourcing channels in order of yield: newsletter writers in your niche, people with a fresh book or product launch, active niche communities, guest-of-guest referrals from people you've already had on, and your own audience. Then run each stranger through a five-part fit score before you spend an hour writing the invite. Source wide, vet hard, reach out narrow.
Most host advice says "invite interesting people" and leaves you staring at an empty calendar. The real problem is two-sided: you run out of people you already know after about episode eight, and you have no way to judge a stranger you found online. This guide fixes both, a map of where strangers worth inviting actually congregate, and a rubric that tells you whether a specific one is worth the outreach before you write a word.
Where do you find podcast guests you don't already know?
Find them through five channels: niche newsletters (the writer or the people they cover), anyone with a launch in the next 90 days, active communities where your topic gets argued out, referrals from past guests, and your own listeners. Each channel surfaces a different kind of person, so working all five keeps your booking pipeline from going stale.
The instinct when the contact list runs dry is to lower the bar, book whoever says yes. Resist it. A guest who is wrong for the show costs you twice: the recording, and the episode that underperforms and drags your averages down. The fix is not "try harder to think of people." It's a repeatable map you can run for an hour and walk away with ten candidates. Here's that map, ranked by how many good guests it tends to return per hour of work.
The five channels, worked one at a time
1. Guest-of-guest referrals (warmest, do this every time)
At the end of every recording, ask: "Who are two people in your world who'd be great on this and would say yes to you?" People you've just had a good conversation with want to be helpful, and they know who's actually worth talking to in their corner. This is the single highest-yield channel, and it costs you one sentence. If you've done ten episodes, you have a referral network of up to twenty warm names you've never asked for.
The reason it works: a referral arrives pre-vetted on the two hardest dimensions, relevance and whether the person can hold a conversation. Your past guest has filtered for both without being asked.
2. Your own audience
The people who already listen are a self-selected pool of topic-obsessed humans, and some of them are practitioners with a story you haven't heard. Put one line in your show notes and a periodic on-mic ask: "Doing something interesting in [topic]? Reply and tell me." This matters more than it used to, because 57% of listeners now rely on social media for podcast recommendations, the first time that source has edged out friends and family at 54% (InsideRadio, citing Coleman Insights/Amplifi Media data). The audience that found you through a clip is reachable through the same channels you're already posting on.
3. Niche newsletters
Every niche has two or three newsletters that the serious people read. Two guests hide in each issue: the writer themselves (they have a point of view and an audience to bring), and the people they quote or link. Subscribe to the five biggest newsletters in your topic and keep a running list of names that show up more than once. Recurring names are a free signal that someone is both active and credible.
4. Launch calendars
Anyone with a book, a product, a study, or a tour landing in the next 90 days is unusually motivated to talk, they need the reach, and they've just organized their thinking into something quotable. Watch publisher catalogs, Product Hunt, Kickstarter in your category, and conference speaker lists. A guest with a launch shows up prepared, which lifts your recording quality and your odds of clippable moments.
5. Niche communities (coldest, but the deepest well)
Subreddits, Discord servers, Slack groups, and the comment sections where your topic gets argued out are full of people who think hard about it in public, but you don't know them, so the vetting bar is highest here. Lurk first. The person who writes the most thoughtful 400-word comment in a heated thread is often a better guest than the person with the biggest follower count, because they can actually explain themselves. This is where you find the fresh voice nobody else has booked yet.
A word on why this work is worth doing at all: the field is mostly graveyard. There are roughly 4.7 million indexed podcasts, but only about one in ten are still actively publishing (demandsage), and shows die early: nearly half never make it past the first three episodes, and most quitters stall between episodes 7 and 25 Amplifi Media. That cliff lines up exactly with when a host's contact list runs dry. Sourcing isn't a nice-to-have. It's what keeps the show alive past the danger zone.
How do you vet a guest you've never met?
Score the stranger on five dimensions, zero to two each, before you write the invite: relevance to your audience, whether they have a fresh story, whether they speak in moments (not monologues), how low-risk they are, and how reachable they are. Anything totaling six or above is worth an hour of outreach. Below six, keep sourcing.
The point of a rubric is to make the cut early, when it's cheap, instead of after you've recorded a flat episode. Researching and booking a guest is a real time cost, you'll want the full process in how to research a guest in under 90 minutes, so the fit score is the gate you run first.
How to read each dimension fast:
- Relevance (0–2). Two if the person's core topic is one your audience would actively choose; one if it's adjacent; zero if you're stretching to justify it. The stretch is how you book a guest who's interesting to you and boring to them.
- Fresh story (0–2). Search their name plus "podcast." If they've been on twenty shows this year, they'll give you the same rehearsed answers everyone else got. A guest who rarely does podcasts but has lived the story is worth more than a circuit regular.
- Speaks in moments (0–2). Watch two minutes of any talk, panel, or existing clip. Can they land a thought in fifteen seconds, or does every answer become a three-minute ramble? This dimension predicts whether the episode produces anything you can post.
- Low-risk (0–2). A ten-minute search for controversy isn't paranoia, it protects your show and your other guests. Zero this out and reconsider entirely if something serious surfaces.
- Reachable (0–2). A brilliant guest with no path to their inbox scores low for a reason: an invite you can't deliver is worth nothing. Two if you have a referrer or a real email; zero if it's a locked-down celebrity with a publicist wall.
Common mistakes when sourcing guests
Most sourcing problems aren't about effort. They're about chasing the wrong signal.
- Chasing follower count over fit. A 200K-follower guest who can't speak in moments gives you a flat episode and a polite no on promotion. Score relevance and "speaks in moments" above reach. Audience size is the last tiebreaker, not the first filter.
- Booking the circuit regular. The person who's been on forty shows this year is easy to book and gives you the answers forty other hosts already got. Fresh beats famous. That's the whole point of sourcing strangers instead of recycling the obvious names.
- Skipping the ten-minute risk check. One unvetted guest with a buried controversy can cost you sponsors and other guests' trust. The low-risk dimension exists for this. Never zero it out to save time.
- Pitching before you can deliver. Falling for a dream guest with no reachable inbox burns days. Score "reachable" honestly, or line up a referrer first through the guest-of-guest channel.
- Sourcing once and stopping. The pipeline goes dry in a month if you don't keep it filled. Block 30 minutes a week to add five names. Consistency in sourcing is what makes consistency in publishing possible, and consistency is the survival metric Amplifi Media.
Turn a found guest into a great episode
Sourcing gets the right stranger in the chair. The interview is a separate skill. Before you record, read the guest properly, start with how to research a guest in under 90 minutes, and build a few questions that get past their rehearsed lines, covered in how to ask questions that get real answers. On the day, the follow-up question is where the episode lives, and that only works if you're actively listening instead of waiting to talk.
One more thing that makes strangers say yes: make the experience easy and respectful. A host who runs a tight, comfortable recording gets referrals and re-bookings, see how to make guests feel comfortable on mic, and know what the etiquette rules guests wish hosts understood look like from the other side.
FAQ
Where do you find podcast guests for free?
All five sourcing channels are free: ask past guests for referrals, ask your own audience in show notes, read niche newsletters, watch launch calendars (Product Hunt, publisher catalogs, conference speaker lists), and lurk in niche communities. Paid guest-matching databases exist, but referrals and newsletters out-yield them for most shows and cost nothing but time.
How do you find guests for a brand-new podcast with no audience?
Lead with the guest-of-guest and community channels, since you can't yet ask an audience you don't have. For your first ten episodes, invite people who are building their own following and want the reach, a guest with a launch in the next 90 days is unusually likely to say yes to a small show because they need every channel they can get.
How do you vet a podcast guest you don't know?
Run the five-part fit score: relevance, fresh story, whether they speak in moments, how low-risk they are, and how reachable they are, zero to two each. A total of six or higher is worth an outreach. The ten-minute risk search is non-negotiable; one unvetted guest can cost you sponsors and other guests' trust.
How many guest candidates should you keep in your pipeline?
Keep enough to cover four to six episodes ahead, and add five new names a week. Running dry is what forces hosts to lower the fit bar or skip a week, and shows tend to die early, with nearly half never passing the first three episodes and most quitters stalling between episodes 7 and 25 Amplifi Media. A standing pipeline is cheap insurance against podfade.
Should you pay podcast guests or use a paid matching service?
For most independent shows, no. Referrals, newsletters, and communities surface higher-fit guests than pay-to-match marketplaces, which tend to over-index on people who'll appear anywhere. Spend the money on production instead. The exception is a tight-niche show where the right experts are genuinely hard to reach without an intermediary.