How to transcribe a podcast episode

Most people transcribing a podcast episode aren't doing it for themselves — they already listened to it. The transcript is for the show: search engines that can't parse audio, listeners who'd rather skim than listen, and future episodes or blog posts that can be repurposed from it.

Why a raw transcript is often the wrong deliverable

A word-for-word transcript is exactly right for accessibility and archiving — nothing should be edited out of a record meant to represent what was actually said. For show notes or an episode page meant for search and for readers, it's usually a starting point, not the finished thing. A wall of unedited speech — filler words, false starts, run-on sentences — reads fine as audio and poorly as a page of text. Breaking it into paragraphs by topic and trimming the verbal filler takes a few minutes and makes the difference between a page people bounce off of and one they actually read.

The SEO case is worth being explicit about: a search engine indexes text, not audio. An episode with no transcript is invisible to search regardless of how good the content is: the transcript is what actually gets crawled and ranked.

Podcast Transcription takes the episode as audio or video. For an episode over the free 90-minute cap, running it through Audio Summarizer first is a fast way to get the gist before deciding whether the full transcript is worth the extra trimming step.