Your Podcast Tour Has 40 Episodes — Your Website Has Zero Pages About It
A consultant we talked to last year had done 43 podcast appearances in about 30 months. Business, marketing, and niche-industry shows, a few with real audiences — one host claims 80,000 downloads an episode. He'd talked for 45 minutes to an hour each time, unscripted, about the exact problems his consulting practice solves. That's roughly 35 hours of him explaining his methodology, answering hard questions, and establishing himself as the person who knows this space cold.
His website's "Media" page had 43 logos. Podcast artwork, arranged in a grid, each one linking out to the host's website. No summaries. No transcripts. No indication of what he actually said. If you searched his name plus any topic he'd covered — pricing strategy, client retention, whatever the episode was actually about — nothing from his own site showed up. The only proof that conversation ever happened lived on someone else's domain, in an audio file Google can't read and an LLM can't cite.
That's not a media page. That's a tombstone with a list of names on it.
Why the Authority Accrues to the Host, Not to You
Here's the mechanism, and it's not complicated once you see it: search engines and AI models build authority around entities — a person, a brand, a claim — by finding that entity mentioned consistently, in text, across multiple sources that reinforce each other. When you appear on a podcast, the host's site gets an episode page with show notes, a title, maybe a transcript. That page is indexed. It has your name in it. But it lives on their domain, targeting their audience, optimized for their keywords, feeding their authority — not yours.
Your own site, meanwhile, has a logo. A logo has no text for Google to crawl. It has no claims, no quotes, no topic association. So when someone searches "[your name] client retention" or when ChatGPT gets asked "who's a credible voice on B2B pricing strategy," the model has 43 opportunities to cite you and instead finds one grid of images that says nothing. You did the work. The host's site banked the SEO equity.
This is the same failure mode as the press-mention problem — where a feature links to a landing page instead of the person — except worse, because at least a press mention exists as indexable text somewhere. A podcast appearance without an owned page doesn't exist anywhere on the searchable internet except as 45 minutes of audio nobody transcribes and nobody reads.
Audio Isn't Indexable — Your Expertise Has to Be Rewritten in Text to Exist
Google can technically transcribe audio through YouTube's captioning, and some podcast hosts do publish transcripts. But even when a transcript exists, it's raw and unstructured — a wall of "um" and cross-talk that no algorithm treats as authoritative content, and no human reads. Answer engines like Perplexity and Google's AI Overviews are pulling from clean, well-structured, citable text: a paragraph that states a claim, attributes it, and lets the model quote it back with confidence. A messy verbatim transcript on someone else's site doesn't qualify. Neither does silence on your own.
This is the part most personal brands miss when they think about the site they were told they needed. The pitch is always "you need a professional website" — and they build one, with a bio, some services, a contact form. What nobody tells them is that a personal brand website has a second job beyond looking credible to a human visitor: it has to be the canonical source that search engines and AI models point to every time your expertise gets referenced anywhere else on the internet. Every podcast, every guest post, every conference panel is a data point that should route back to your domain. If your domain has nothing to route to, the data point evaporates.
The Fix Is an Appearance Page, Not a Logo Wall
The architecture that actually works isn't complicated, but almost nobody builds it because it requires treating each appearance as content production, not just calendar management. Every podcast, panel, or guest feature gets its own URL. Not a shared "Media" page with a grid — an individual page, indexed separately, built around what was actually said.
That page needs four things to carry real weight. First, the embed or link to the original episode, because you want the reciprocal signal and you're not hiding that it happened elsewhere. Second, a 400–600 word summary written in your voice, not a copy-paste of the host's show notes, covering the two or three specific claims or frameworks you talked about — the actual insight, not "we had a great conversation about marketing." Third, pull quotes: three or four sentences lifted directly from what you said, formatted as blockquotes, because these are exactly the citable units an AI answer engine is built to extract and attribute. Fourth, structured markup — PodcastEpisode or Article schema with author set to you — so the page is machine-legible as a specific claim made by a specific named person, not just a blog post.
What Forty-Three Pages Do That One Logo Grid Never Will
Multiply that by 43 episodes and you don't have a media page anymore — you have 43 indexed pages, each targeting a different long-tail topic you're already an expert on, each one reinforcing the same entity (you), each one capable of being surfaced independently when someone searches that specific question. One page might rank for "how to price a retainer without scaring off a client." Another for "when to fire a client." You didn't write 43 blog posts from scratch — you already did the talking. You're just capturing it instead of donating it to someone else's domain.
This also solves a distribution problem nobody talks about: podcast audiences are one-time. Someone hears you on an episode in March, and unless they subscribe to your newsletter on the spot, that exposure is gone. An indexed appearance page keeps working in July, in October, in a Google search two years from now, in an AI Overview that has no idea when the original conversation happened.
The Compounding Problem You're Not Solving
Every quarter you go without this, the gap gets more expensive, not less — because the appearances keep happening and the debt keeps compounding. A consultant who's done 43 podcasts and captured none of them isn't just missing 43 pages. They're missing the pattern-recognition effect that happens when a search engine or an AI model sees the same entity making structurally similar, reinforcing claims across dozens of indexed pages over years. That pattern is what makes a model confident enough to cite you unprompted. One page won't do it. Five might start to. Forty-three, done right, is the difference between being a guest and being a source.
The uncomfortable question worth asking this week: pull up your own site and count how many of your public appearances — podcasts, panels, guest posts, quoted mentions — have an actual page on your own domain. If the number is zero, or close to it, you've been building someone else's authority for years and calling it marketing.
If you're sitting on a backlog of appearances and no architecture to capture them, that's a conversation worth having before the next podcast booking, not after the fortieth one. We'd rather talk through what your site should actually be capturing than watch another quarter of proof walk out the door to somebody else's domain.