Technology for transcription when time and budgets are tight20.07.2018
Have you ever stopped to ask yourself, “Am I doing this in the most efficient way? Could I be doing this smarter or faster?”
At HRW, we take efficiency seriously and at the core of this ethos is ensuring that what we do produces desirable results whilst making the most of the materials, time and energy we have at our disposal.
We are always discussing ways to innovate as a team at HRW, whether it be updated takes on old classics, or new territory completely. For this reason, we recently hosted our very own ‘HRW Innovation Challenge’ – a creative competition, open to the whole company, to submit ideas big and small that they think have the potential to add value to our business and our clients. It’s a broad challenge and ideas can come from many different areas across the business, such as new research methodologies, techniques, internal processes, or improved ways of working with partners. Many of the submissions were simple but highly effective ideas; looking at how we could elevate the way we work, how we carry out everyday tasks.
Forming a passionate team, Elizabeth Openshaw, Emily Glass and I set about racking our brains for inspiration and an idea to submit and pitch to the judges. We put our heads together and thought about areas and tasks within our everyday profession that we felt could be enhanced.
Working in the world of healthcare market research, moderating and analysing interviews with different types of stakeholders is a large part of our role. Whilst carrying this out, we are always trying to work as efficiently as possible – in saving our own time and money, but most importantly, saving our clients’ budgets. On most projects, we have a local language person type up transcripts or content analysis for all interviews. Transcripts or secondary analysis then allows us to fully concentrate on our discussion and objectives during the interview, rather than having to take notes as we go, or type them up afterwards.
However, as in all businesses, we can be faced with unexpected budget cuts or time pressures that don’t allow for transcription, and in these scenarios sometimes we have to make do without such analysis tools. It can be challenging for any moderator to take notes whilst interviewing and listening back to recordings can be incredibly time consuming. We sought a crafty solution that would allow us to be effective in such situations, to deliver rich insights and quotes from the research even when time and budgets are tight. With this in mind, we entered the world of voice recognition technology and specifically, automated transcription services.
Voice recognition technology promises to take in raw audio files (such as interview or meeting recordings) and deliver a transcript of what was said. After reading some reviews online, each member of our team chose a provider of this service (Windows, Google and Dragon) to ‘road-test’ and feedback on. We found several pros and cons for each including benefits such as free subscription, no installation requirements, and easy accessibility. However, drawbacks included wide-ranging accuracy (especially for fast talkers or strong accents), lack of punctuation, and unsuitability for larger groups such as focus groups where the software has difficulty determining different voices.
We concluded that, dependent on needs from your voice recognition technology, there were a couple of options to suit. The free platforms with more varied accuracy are more suitable for supporting informal, personal note taking and creating shorter transcripts where needed. The more established platforms (some of them incurring a monthly cost), are better suited to use on the move when you need something more reliable and accurate, such as clearer interview notes or recording meetings. All in all, using these types of platforms gives us instant access to written up interviews or meeting notes, which in a time-pressured professional environment can be very valuable!
Our success so far gave us inspiration to look at other offerings available on the web. Subsequently, we found a plethora of options – all with their own shortcomings and advantages, but dependent on your needs – there is something to suit most. There are possibilities for further, detailed analysis, such as integration of audio/video and transcript text allowing you to quickly search for key works or questions in an audio to get a quote or clip. We often utilise methodologies such as short 10 minute follow up longitudinal resonance interviews or video / audio diaries, and these types of platforms can be very useful in helping to analyse at lower costs and quicker timeframes.
All in all, automated transcription has found its place among the digital tools we’re able to put to good use where needed, allowing us to be more efficient in our analysis, therefore better able to serve our clients in a cost and time pressured environment.
By Jess Woodhead