Audio to text (also called speech to text), is the ability to retrieve text data from an audio/spoken language source.

This allows us to detect mentions of brands, products, campaigns, etc. even when they are part of the audio and are not included in the text/caption of the post; increasing our ability to detect mentions relevant for your brand.

Audio to text Availability on Launchmetrics platform

Audio to text is available for YouTube, TikTok, Red and Douyin.

For Tiktok and Youtube, we rely on the subtitle track of the platforms itself. Thus audio to text may not always be available.

For Red and Douyin, we use Aliyun transcripts API. Please note that for Douyin, audio to text availability is restricted to posts with a duration of less than 30 minutes.

The audio to text is available for TikTok documents monitored after May 8, 2023, for YouTube documents monitored after June 7, 2023, for Douyin documents monitored after October, 2021and for Red documents monitored after October, 2022.

Languages

The Audio to text information does not have any language restriction, as long as it is available in the source, it will be available in the Launchmetrics platform, regardless of the language.

Accuracy

The audio to text sources are sometimes auto generated or user generated (subtitles track), as such they may contain typos or not be fully accurate.

However, the results have been validated and the quality is consistently good.  We have measured the accuracy of audio to text data at 93%.