Glossary
What is a transcript?
By the KeyMo team · Updated
A transcript is a written copy of everything said in a recording, in order, often with speaker names and timestamps. For a podcast or video it is the searchable text version of the audio. It differs from captions, which are timed to appear on screen, and from show notes, which only summarise.
Transcript vs caption vs show notes
| Text | What it holds | Timed to the video? | Made for |
|---|---|---|---|
| Transcript | Every word spoken, plus important sounds | Optional timestamps | Reading, search, accessibility |
| Caption | Speech and key sounds in short lines | Yes, line by line | Watching with sound off |
| Show notes | A summary, links and guests | No | Choosing whether to listen |
The W3C accessibility guidance says captions and transcripts hold the same text, so one can be used to make the other. A basic transcript covers speech and non-speech audio. A descriptive transcript adds a text description of the visuals needed to understand the video.
Why a transcript drives clip finding
Reading is faster than listening. With a transcript of a one-hour episode you can search for a phrase, a name or a question, then jump to it. Strong clips often contain a clear claim, a story or a contrast, and those show up when you skim text. Timestamps then tell you where to cut.
The weak point is accuracy. The W3C guidance says automatically generated text does not meet accessibility needs unless it is confirmed fully accurate, and that missing a single word such as "not" can flip the meaning. Read before you quote.
Do I have to publish one?
The W3C guidance says transcripts are required at WCAG Level A for pre-recorded audio-only content, such as a podcast. That is an accessibility standard, not legal advice, and local law varies. Many shows publish transcripts anyway so listeners and search engines can read them.
Common mistakes
- Publishing machine text unchecked, with misheard names and numbers.
- Leaving out speaker names in a conversation, so nobody can tell who said what.
- Pasting a raw transcript as show notes. Notes should help a reader decide; a transcript should help them search.
- Treating a caption file as the full record. It may be trimmed to fit short lines.
Where KeyMo touches it
KeyMo does the reading for you. It scores moments in your video out of 10, keeps only those at 7 or above, and gives each one a line of reasoning, so a weak episode returns fewer clips. Its captions run word by word. If you work from text yourself, our transcript cleaner tidies a raw copy.
Sources: W3C WAI, Transcripts in Making Audio and Video Media Accessible (checked 11 October 2026); W3C WAI, Captions/Subtitles (checked 11 October 2026); W3C WAI, Transcribing audio and video for captions (checked 11 October 2026).