What is speaker identification?
Updated Oct 25, 2026
Definition
Speaker identification is the part of transcription that works out who said each part of a meeting, so every line of the transcript is labelled with a person's name instead of one block of text.
Why it matters
- Action items and decisions go to the right person.
- You can follow one person's points through a long meeting.
- Quotes in summaries and reports can be checked against who really said them.
What affects accuracy
Correct display names in the meeting, one person speaking at a time, and each person on their own microphone. People sharing one laptop in a meeting room are the hardest case.
How Acta names speakers
Acta labels each line of the transcript with the speaker's name, taken from the meeting and from the attendees of your calendar invite. When a recording gives no names, people show as “Speaker A”, “Speaker B” and so on, rather than a wrong guess.
Meeting transcription helpFAQ
What's the difference between speaker identification and diarization?
Diarization splits the audio into different voices (Speaker A, Speaker B). Identification goes one step further and puts a real name on each voice.
Why does my transcript say “Speaker A”?
The recording didn't come with a name for that person, so Acta labels them by voice instead of guessing. The words are still split by speaker.