Who are the users of automated transcription?
During the last year we have witnessed a huge increase in the use of automated transcription services. Although the quality of automated transcription is not yet perfect, it is already useful for many purposes.
The use cases of transcription and the required accuracy differ from each other. In some cases, accuracy reaching 100 % is essential, but for other situations smaller accuracy is enough. In this post we will discuss more closely the most potential early adopters of automated transcription services.
Personal use right after the interview
In many professions the transcriptions are used solely for own use, which is why the demands for accuracy are not that high. In case the person utilizing the transcript has also participated in the transcribed discussion, he is able to fill the potential blanks if speech recognition technology has done any mistakes.
This is applicable for example within the media sector where journalists conduct the interview and write the article e.g. during the same day. In this case the automated transcript is used for “refreshing the memory”.
Qualitative research conducted e.g. in the university is largely based on interviews, and they create a big part of the data collection. Requirements for automated transcript accuracy are not that strict in case the researcher is personally conducting the interviews.
One common use case for automated transcription is podcast transcription. Podcast producers want to focus on creating high-quality podcasts rather than to spending their time to transcription.
In many cases the automated transcript can be good enough for the podcast producer. In podcasts the main purpose of the text is often to increase search engine visibility and give their followers another format where they can easily scan through the podcast.
Automated transcription is way more cost-efficient solution than human-made transcription. Usually human-made transcriptions cost around 80 $ / audio hour, whereas the price of automated transcription can be as low as 5 $ / audio hour. That makes automated transcription 16 times cheaper than human-made transcription!
This lower pricing opens entirely new markets where transcription services can be utilized. Thanks to this, different kind of companies can e.g. transcribe all their important meetings to text format without huge costs.
For students who need to do transcription for their school work can now make use of automated solutions instead of doing the hard work manually themselves.
In this post we have referred several times to the accuracy of automated transcription regarding the general level. In reality, the accuracy of each audio is different, and the usability of some files might vary a lot.
If you are willing to know how to increase the likelihood of getting good accuracy transcript, check out this post from our blog.