Publish text that represents the conversation.

Choose an editorial convention

Decide how your publication handles fillers, repeated starts, overlapping speech, and unclear words. Keep the convention consistent without polishing a guest’s answer into a stronger claim. Label substantial condensation as a summary rather than calling it a full transcript. Use approved speaker names, not identities inferred from voices.

Review against the right recording

Work from the release cut and record its version. Check speaker transitions, names, numbers, and qualifiers before punctuation. Keep questions for the producer when a passage remains unresolved. A well-formatted draft can still be wrong, so visual polish should not determine whether it passes review.

Make the published text useful

W3C guidance explains transcripts as text representations of meaningful audio information and discusses readable presentation. Decide who approves the finished page and how readers report errors. When an important correction is made, review related summaries and quotations as well as the transcript so the old mistake does not persist in derivative copy.