ra2 studio - Fotolia
Microsoft launches Word transcription tool
The new Microsoft Word transcription feature makes crafting notes from an audio file a more streamlined process. The service is available with a Microsoft 365 subscription.
Microsoft has launched a Word transcription tool that could provide some companies with an alternative to third-party services.
The Transcribe in Word feature lets users upload audio files to the Word online web app, which creates a written record of the meeting, interview or conversation. The tool, launched last month, also identifies different people when they speak.
The Word transcription feature should be attractive to students, reporters and office workers who want a searchable version of a recording, industry observers said. It competes with third-party offerings such as Otter.ai.
Microsoft is capitalizing on its market-leading word processor to gain an advantage over competitors. Transcribe in Word is included in a Microsoft 365 subscription at no additional cost.
The company's new feature is like other speech-recognition tools, in that it doesn't provide an error-free record of a webinar or discussion. To significantly reduce mistakes, an organization would have to use a human to transcribe the recording.
"There are limits to this thing," Gartner analyst Craig Roth said. "I don't see it necessarily upending a lot of business models."
Currently, the feature works only with the Word web app. It also limits each user to five hours of recordings a month. English is the only supported language.
Replacing clunkier options
Before this tool, Word had a dictation feature that transcribed live speech. People trying to use that service in place of an audio-to-text service likely had poor results, Roth said. He recalled when he played an audio file into his computer's microphone, an awkward method of achieving the result promised by the new Word transcription feature.
"At that point, I was essentially using the dictate feature [for] transcription, just in a very cheesy way," he said. "When I was doing that, sometimes it would work beautifully, and other times it would choke halfway through, and I would have to try again."
In contrast, the Word feature is a streamlined process for creating a transcript from various audio files, including .mp3, .wav, .m4a and .mp4. After a user uploads an audio file, it provides a transcript in a sidebar to a Word document. Time-stamped audio attached to the transcript allows users to review what was said to correct errors.
While transcription tools are useful, they are also limited, Roth said. When people speak, they are more informal and make mistakes that listeners compensate for -- writing requires a more careful approach. As such, this feature is better suited for creating easy-to-review personal notes than crafting a formal document, he said.
Microsoft said it is using AI technology in its Azure cloud to deliver Transcribe in Word. It has also been using Azure AI to bolster other products within its 365 productivity suite, including PowerPoint, Excel and Outlook. Earlier this year, the company added a "Presenter Coach" to PowerPoint that used AI. The feature determined whether a presenter was talking too fast, speaking in a monotone or using too many filler words.
