What is the difference between dictation and transcription?

Dictation writes text as you speak it. Transcription writes down speech that someone already said and recorded. The timing, the author and the purpose all differ. In dictation you are the author, you speak the words you mean to write, and you correct them as you go. In transcription the audio comes first, the speakers may be several people, and the output is a record of what was said. The what is dictation guide introduces both terms.

Dictation writes your words as you say them

The dictionary sense of dictation already points at the difference. Merriam-Webster defines dictation as the act or manner of uttering words to be transcribed ¹. The Cambridge Business English Dictionary describes it as speaking aloud for a person or machine to record ². Those older definitions treat dictation as the speaking step, with someone else writing the words down later.

Modern dictation merges the two steps. Voice typing on a computer or phone takes the words as you speak and types them where your cursor is. The author is still you, and you see the text while you can still fix it. A dictation tool does this live, while a transcription tool works from audio that already exists.

Transcription turns a recording into a record

Cambridge gives transcription two senses: a written record of words or music, and the process of transcribing something ³. In practice, transcription starts with audio or video: a meeting, an interview, a lecture or a voicemail. It can have many speakers, and it is judged by whether it matches what was said.

Companies sell transcription as a service. Microsoft lists transcribing calls in real time or in batches among the uses of its Azure Speech service ⁴. Phones can do it too. On an iPhone, a Voice Memos recording can be turned into a transcript from the recording's own screen ⁵.

Side by side: the same speech goes two different ways

The table shows the usual pattern. Dictation is about writing, transcription is about recording, and the editor's job is different in each.

Why the difference matters: errors land in different places

Speech recognition makes mistakes in both jobs, and the fix depends on who catches them. A dictated draft is checked by its author, who knows what was meant. A transcript is often checked by someone who was not in the room.

Medicine shows the gap. In clinical notes dictated in 2016, the speech recognition draft had a 7.4 percent error rate, or 7.4 errors per 100 words ⁶. Medical teams have long put a transcriptionist between the recording and the signed report. The US Bureau of Labor Statistics describes that work as listening to the recorded dictation of physicians and transcribing it into medical reports ⁷.

It depends: the same audio can be either job

The label depends on who is writing and when:

• You speak notes into a recorder, then type them later. You did the speaking, and someone else did the writing. That is dictation in the older sense, and the typing is transcription.
• You dictate a memo live, while watching the screen. That is dictation. You are the author, and the text is what you meant to write.
• You attend a meeting and a tool writes it down. That is transcription, even though you were present. You did not write the text.
• You record yourself and transcribe it. The recording exists first, so the output is a transcript, even though the words are your own.

What to do: name the output before you pick a tool

1. Decide whether you are writing or recording. If the text is something you want to write, you need dictation. If it is a record of speech that already exists, you need transcription.
2. For dictation, use the tool in the app where the words go. The how to turn on voice typing guide covers the setting on each device.
3. For transcription, start with the recording. The transcribe audio to text guide covers the options, and the free audio to text tool works in your browser.
4. Proofread the result either way. The proofreading question explains how to find the errors that speech recognition leaves behind.

The related questions are is speech to text the same as dictation, which covers the wider category, and dictation vs voice control, which separates writing from running the computer. The transcriptionist entry defines the medical job, and the transcription overview covers the subject. For the full topic, see the dictation overview.