Our business was started with podcasting, and as such our transcripts tend to be a bit media focused. The format is interview friendly and has some interesting defaults - such as just marking advertisements, promos and songs (e.g.: [ads]), instead of transcribing them. Another noteworthy section of our style guide involves audio with a large number of undifferentiated speakers - in that case we'll start to refer to the by a collective label, for instance in a Q&A section, we'll label the all speakers from the audience as "Audience".
Another media focused bit is that we specialize in non-verbatim transcription - meaning we clean up filler words. Verbatim transcription can be requested if needed - please mail firstname.lastname@example.org, for verbatim requests.
As a matter of face most of these defaults can be changed by request - usually by leaving a comment for the transcriber.
The finished transcriptions are delivered in plain text, HTML and Word. Additional formats are available in request, in particular SRT captions are availble for any timestamped audio. All transcripts can be downloaded from the Customer Center.