That Product Studio worked with L10GO to create a new closed caption service.
CharlieCC is a new startup aimed at improving the speed and cost of closed captions. Our hybrid solution of Automatic Speech Recognition supplemented by skilled linguist reviews means you'll never pay over the odds for quality subtitles again.
The team size was minimal for this project as we were bootstrapping everything to validate the business idea quickly.
We ran a 2-week lab sprint to quickly research the market and validate the need for a new captioning service.
We tasked ourselves to find as much insight as possible. The team ran workshops and interviews with SME's and accessibility experts. We also did extensive competitive analysis.
80%+ Total web traffic is video content. The sheer volume of video content is set to increase exponentially with 20 billion devices now connected to the internet. 466 million people have a disabling hearing issue. That’s 5% of the population of planet earth.
"We are becoming more socially aware of a wider range of physical and mental health issues. Now that we are talking much more openly about disability, more people are fighting to make sure services are more inclusive. Law changes from the FCC in America and the EU passing the accessibility bill show us that it is no longer an afterthought."
Next, we put together a Business Model Canvas; this was to see if there were any implications in the business planning process.
We also mapped out our assumptions and ran online brand strategy exercises in Miro to come up with an art direction and tone of voice.
With the insights and assumptions, it was time to ideate and decided on the solution.
We started by doing some Crazy 8's followed by Solution Sketching. We then had a hangout to present the solution sketches and did a round of dot voting. I the end, we used a decision matrix to work out the best way to build the MVP, as some features were already proving more technically challenging in the development spike. We decided on a raw file management CMS and a caption editor that would be able to spit out ASR captions from AWS.
With the rough idea decided, I started to map out some of the flows and architecture of the prototype first.
I then created a basic clickable prototype so we could use it as an artefact to validate, communicate and iterate design decisions.
Throughout the creation of the prototype, we ran Stakeholder and Technical Review to secure buy-in from the team.
As we were rapidly generating this product, we needed to get feedback ASAP. So we recruited transcribers to do cognitive walkthroughs and gather feedback from the initial prototype. We did multiple rounds of usability testing throughout development, ensuring we implemented the best possible UI for fast transcription.
"Is there a shortcut for that?"
As most transcribers on the platform are pros when it comes to typing, 99% of the human-computer interaction is via their keyboard. So we needed to make the editor fully functional without the use of a mouse. This was a fun challenge, and we managed to increase the caption per minute time by 20%.
CharlieCC is live, it has already accumulated over ten clients, 40 transcribers and is turning around over 5,000 video minutes a month. You can find out more about the final product here: http://charliecc.co.uk
