When you’re trying to create something that you want people to buy you must test that it actually works, right? That’s just plain logical.
For NewsLing, we put off our logical logistics testing for far too long. We had all these hypotheticals and theories as to how we would make our product work and no hard tested data or feedback from credible sources.
Our product, NewsLing, is a service that takes a local newscast and translates it into Spanish. From our research and user testing, we recognized there is a great need for local news in Spanish, and with NewsLing we are trying to address this problem.
Our current prototype is geared towards the end-user. How we made that prototype does not reflect the process we outlined in our business plan that we would pitch to our customer, a broadcast news station.
The steps for getting a live translation of a newscast goes like this:
- One hour before the program we give the two simultaneous interpreters the most updated script of the broadcast that the station can provide.
- The interpreters get one hour to go through the script to prepare for the broadcast.
- The interpreters get set up in a sound booth to record and stream the audio through the app.
- They translate the 30 minute broadcast
- They break and prepare for the next one.
This was the process and we put it in a smaller test. Instead of a full hour to look over the script, the interpreters had 30 minutes. We set up the school’s sound booth with headphones, the video of a WRAL broadcast and recording equipment the translators would. We contracted two representatives from CHicle, Amanda Black and Perla Saitz, to help us in our experiment.
They came in at noon, looked over the script for 30 minutes and stepped into the sound booth to simultaneously translate the audio of the broadcast, After an hour and 15 minutes, they were gone.
Did our process crash and burn? Nope.
Here is what we observed:
When looking over the script, they talked through it to hear what it would sound like out loud. They looked like they needed a reference tool; anything like a Spanish-English dictionary or access to wordreference.com. They went through the script meticulously but didn’t get through the whole thing.
In the booth, they started strong and were prepared to tackle the stories they had reviewed and the order they had come in. There were moments of silence when the stories and transitions were too fast, the order of stories was changed up or they didn’t know who should be translating.
The weather portion of the broadcast went surprisingly well with some hiccups concerning meteorologist terms.
Thirty minutes is not enough to look over a long script. An hour is needed. They had recommendations to include a glossary with each topic of the story to refer to as they translated. They said the script was useful as a good preparation for the live translation but distracting while actually recording. The translators also included a $40,000 salary as a reasonable estimate for a consistent work schedule.
Our process had a solid run through but will have to be adapted with some of the recommendations by the translators.
But, IT’S DOABLE!!!!!!
I think we were worried that if we tested it and it didn’t work, all of our reasoning and excitement on the user testing side would go unanswered and unfulfilled.
It is a scary thought and if we had found it didn’t work, we would have had to work that much harder in finding another way to do it or explaining in a very detailed way how and why our idea didn’t work. Either way we would have learned a boatload on the value of testing the process and logistics of a business.
Happy Dance until we remember how much it costs for the entire process.