The transcript passes through three pairs of hands
First the reporter writes it — keystroke by keystroke on a stenotype machine, with audio running as backup. Then a scopist reads the reporter's untranslated notes against that audio, resolving conflicts, researching spellings, and flagging anything the record cannot support. Only then does the reporter proofread, certify, and sign. A transcript you receive has been read end to end at least twice by people whose job is finding what is wrong with it.
What scoping actually catches
Working reporters describe the failure it prevents: a long transcript full of mistranslations that someone has to recover line by line — untranslates the software guessed at, proper names rendered phonetically, timestamps that drift from the audio. A good scopist returns the reporter's own writing to them as questions, not corrections: this stroke could be two different words, which one did the witness say? The reporter's answer, checked against the audio, is what makes the certified version certain.
Why the reporter still owns every word
Scoping is preparation, not authorship. The certificate at the back of the transcript carries one name, and that person answers for the record. That is why turnaround tiers exist: standard delivery buys the full read-through; expedites compress it with more scopists, not fewer; and a rough draft skips it openly, stamped uncertified on every page.
What to ask when comparing agencies
Ask who scopes, whether the scopist works from the reporter's audio or from text alone, and what happens when the notes and the audio disagree. An agency that cannot describe its own quality pipeline is telling you it does not have one. Ours runs every job through the same bench — see the turnaround tiers — and the reporter's name is on the certification page either way.