Save Time With Microsoft Word’s Built-In Transcription Feature

Thanks to a powerful new productivity enhancement, Microsoft 365 subscribers can now transcribe audio files with multiple speakers, at no extra cost. Our tips and tricks show you how to save time and money with Transcribe.

By Jill Duffy

Microsoft recently announced a new feature in Word that automatically transcribes audio files, specifically conversations between two or more speakers. For people who regularly record interviews, talks, and other conversations, this new transcription tool lets them get written text of their audio files with very little work and no extra cost, which is all a productivity boost.

I gave the new feature a spin and have to admit that even with a few first-timer problems, it saved me between one and two hours of time for a 22-minute recording. Plus, thanks to the way the feature is implemented, it’s easy to listen to the audio file and make corrections to the text at the same time. I found Word’s transcription feature to be an incredibly useful time-saver. Here’s what you need to know about it, how to get it, and how to use it.

How to Get Transcribe for Word

Transcribe for Word is currently available in the web version of Microsoft Word, and you have to be a Microsoft 365 subscriber to get it. You also must use the Chrome or Microsoft Edge browser to access this feature. To find the tool, go to the Home toolbar and look on the far right side for Dictate. Click it, and find the Transcribe option from the short drop-down menu (it’s the second choice).

As of this writing, Transcribe is only available for American English, though, in the future, it will work with whatever language you set as your preference in Word. For uploaded audio files, there is a size limit of 4MB. Finally, you need an internet connection for the feature to work.

What Is Transcribe, and How Is It Different From Dictate?

Transcribe is a new feature that transcribes an audio file into text. You can record audio directly in Word or upload an audio file that you record elsewhere.

How is it different from the Dictate feature? Dictation interprets one person speaking. Transcribe handles multiple speakers. Dictate is something you do instead of writing but usually while still thinking about how you will turn your words into a final piece of writing. When you use the Transcribe tool, you likely want to focus on the conversation and save your thoughts about how you will use those words in the final piece for later.

Security and Privacy With Microsoft Transcribe

How does Microsoft turn audio into text, and is it private and secure? According to a press briefing, the company uses AI rather than humans to transcribe the audio. Any file you upload or record gets saved to your OneDrive account. Look for a new folder called Transcribed Files in OneDrive. These files are protected by the same privacy policy as other files you save to OneDrive.

When you share a Word document with someone and it contains a transcribed file, the recipient can see everything in the Transcribe pane. In other words, they can see the full, original transcribed text in text blocks with time stamps. They cannot, however, play the audio file.

How Does Microsoft Transcribe Work?

As mentioned, you can record conversations from Word or upload an audio file to be transcribed. I tested the feature by uploading a 22-minute file, and it took about 30 minutes to transcribe. If you record in real time using Word, however, the app sends pieces of your audio to the AI transcriber as you go, which means you get the final transcription back much faster.

In my test file, and in other demos I watched from Microsoft, the app was pretty good at differentiating voices. My test file had two female speakers. The app initially decided there were three people. It labels people as Speaker 1, Speaker 2, and so on. In the end, the app only misidentified one line as from a third speaker, so it was very easy to correct.

The final transcription appears in a panel to the right of your document, rather than in the document itself. That way, you can choose to include the entire transcription or just certain quotes.

Tips and Tricks for Microsoft Transcribe

Here are a few tips to working productively with the new Transcribe tool.

1. Rename Speaker 1, Speaker 2 First. Before doing any other editing, you should rename Speaker 1, Speaker 2, and so forth in the Transcribe panel. You can do it once and then choose to correct all instances. If you have more than two speakers, you probably should double-check any quotes you use by listening to the audio and making sure the voice lines up with the name, as it can still go wrong.
2. Add Text to Your Document Piece by Piece or All at Once. Each transcribed block of text has a plus sign in its block. When you click that plus sign, it adds the text to your document wherever your cursor is. Alternatively, you can add all the text to your document by choosing the option at the bottom of the panel that says Add all to document.
3. Pay Attention to Time Stamps. Each transcribed block of text has a time stamp on it. Those are helpful when you want to back up the audio to listen again and make sure the AI transcribed the text correctly.
4. Check for Random Capitalization. One of the most common errors I saw was the random capitalization of words. Pay attention as you edit your final copy.
5. Bad Words Get Bleeped. My test transcription contained the phrase “piss pouring” and the transcription came out as “**** pouring.” Depending on how you intend to use your final document, you may want to look out for bleeped words.

Productivity Booster

The next time you have to transcribe an interview or your want a recording and typed transcript of a meeting, you can use Word to do it with no extra cost (as long as you’re a Microsoft 365 subscriber) and a whole lot less effort than doing it by hand.

For groups that transcribe audio in bulk, a transcription service may be a better bet. Some of these services are even free. Office 365 is not free, but for the right user, having a built-in transcription feature that’s so tightly integrated with Word is a game-changer.

Originally published at https://www.pcmag.com.