Episode 3 · Tuesday, 25 August 2026

When ChatGPT Recommends Your Show

The era of the search box is ending as conversational AI assistants become the primary gatekeepers for podcast discovery and listener recommendations.

Conversational SearchPodHerdAI DiscoveryChatGPTGeminiSearchable Transcripts
By How to Get Discovered | 14m listen | 1 chapters
When ChatGPT Recommends Your Show cover
How to Get Discovered · No. 3

About this episode

ChatGPT and Gemini are fundamentally altering the discovery landscape as conversational interfaces replace traditional search boxes. Maya and Tom explain how listeners like the hypothetical Sarah now bypass Google to ask AI assistants for specific podcast recommendations. This shift forces creators to move beyond audio files and focus on the readable text pools that these large language models ingest to determine which shows are worth suggesting.

PodHerd provides a technical solution to this visibility crisis by transforming raw audio into structured, article-like transcripts with AI-generated chapters. Unlike the unformatted machine transcripts that often result in useless grey slabs of text, these structured documents allow AI assistants to cite specific moments within an episode. The Ask the Archive tool further enhances this by acting as a digital staff member that answers listener questions using direct audio citations. Creators can now verify their reach through PodHerd analytics, which track referral traffic specifically from platforms like Perplexity and ChatGPT alongside Google Search Console data.

While AI middlemen risk summarizing content to the point of replacing the original play button, the human element remains a vital safeguard. Maya and Tom argue that a text summary cannot replicate the nervous tone of an expert beekeeper or the unique chemistry of a host conversation. The episode concludes with a preview of The Loyalty Trap, a forthcoming deep dive into why creators should prioritize their existing audience over the constant chase for new listeners.


CHAPTER 06 / 10 Discussion
Show full chapter

PodHerd Structured Transcripts and AI Chapters

PodHerd provides automated transcription, searchable public pages, and AI-generated chapters for podcast archives across all service tiers. The goal is to transform raw audio into a structured article-like format that allows users and AI assistants to navigate to specific moments in an episode.

podherd· transcription services· ai chapters· timestamps· searchable pages

08:32 Transcriptions got much better at the basics, like who's speaking, punctuation, where a paragraph should break. And then you can build on it. Split the episode into chapters, pull out the topics, tie each one to a timestamp so someone can land on the exact moment a thing was said. Now the page reads like an article with a player attached. Go on, say it. Hmm? The name of the company this show's archive lives on. Fine. On every podherd tier, your back catalogue is transcribed and each episode gets a searchable public page with AI chapters and topics.

09:09 Other tools do versions of this too, and the point stands whichever you use. If you're going to do transcripts at all, do them properly. A half-done one is almost worse than nothing because you think the job's finished. That might be the thing we agree on most all episode. Though I'd add one thing. A structured page is only as good as what's on it. If the conversation rambled, you get beautifully organised rambling. Agreed. Chapters won't rescue a dull episode. They just make it easier to find out that it's dull. There's another way to look at all this, though. If people like asking a question and getting an answer, why should that only happen on ChatGPT?