Zoom has become the first web conferencing vendor to offer automated transcription of online meetings -- a feature...
expected to become ubiquitous in the next few years.
Starting in late January, Zoom's business, enterprise and education customers have been able to turn on automated transcription of meetings and webinars through the platform's cloud recording tool. The transcripts indicate who was speaking and include time stamps synced with the audio and video recordings; they are also searchable.
"All of the vendors that are in this market should be paying attention to this," said Adam Preset, research director for the digital workplace at Gartner. "And many of them, I'm sure, are also working on similar capabilities."
Microsoft is previewing a real-time transcription and translation feature in Skype Meeting Broadcast, its webinar platform. And numerous companies offer transcription of meeting recordings for a fee. But Zoom is believed to be the first to offer no-cost automated transcription services to its subscribers.
Preset said Zoom's transcription feature was "a significant differentiator" in the web conferencing market today. But within a few years, he expects rivals, such as GoToMeeting, Fuze, Cisco WebEx and Microsoft Skype for Business, to offer the same type of service. In the meantime, enterprises can continue using platforms such as Amazon Lex, IBM Watson and Google Cloud Speech API to convert speech to text.
"It won't be a long time," Preset said. "It might just be that some enterprises need to be a little bit patient and keep an eye on the solutions that they currently have in place to get what they want eventually."
Automated transcription services have widespread appeal
Automated transcription services have the potential to enhance productivity and save time in a wide range of use cases, according to Roopam Jain, an analyst at Frost & Sullivan, based in Santa Clara, Calif. They should eliminate the hassle of notetaking and make it easier for employees that missed a meeting to get caught up.
"Recordings and playback, while being useful, can be a time-consuming option for many use cases. More so because all the contents of the meeting recording may not be relevant to each user," Jain said. "Recording transcripts allow users to see clearly who is speaking, when they spoke, and do an intelligent search by keywords."
The feature could be particularly useful for industries mandated to keep exhaustive records, such as hospitals and law offices. Many other enterprises could take advantage of automated transcripts to build training libraries or overcome language barriers within a global workforce.
But speech recognition software typically struggles to account for variations in accents, diction and volume, and often produces transcripts "riddled with inaccuracies," Jain said. Zoom is offering its transcription in partnership with AISense, whose machine learning technology could improve the precision of transcripts over time.
Zoom claims its transcription feature is 90% accurate under "ideal" conditions and up to 92.34% accurate with high-quality audio recordings, according to another Frost & Sullivan analyst, Rob Arnold. "Pretty strong," he said.
Automated transcription follows a pattern among web conferencing vendors
Web conferencing vendors are now focused on "the meeting life cycle," ensuring materials, records and real-time communication tools are available before, during and after meetings, Arnold said.
"So, the meeting itself is not just a finite time period, where we go to this one virtual place and meet for a half-hour, and everyone walks away with their own notes," Arnold said. With Zoom's automated transcription services, "everybody walks away with the same notes, and you can stay on the same page."
The success or failure of Zoom's transcription feature could determine how quickly competing vendors bring their own versions to market.
"Once Zoom proves that it is winning in certain verticals or workgroups because of this feature, this could cause others to add it," Arnold said.