Over the course of several months in 2015, I worked with the Pop Up Archive team to rapidly prototype and test a series of features to showcase the Audiosear.ch API, including PodLikeThat, Pod-A-Day, Tastemaker Picks, and Audiosear.ch show search filtering.
Wireframing, Card-sorting, A/B Testing
Sketch, Optimizely, HTML/CSS, Mixpanel, Google Analytics, Paper Prototyping
At Pop Up Archive, we regularly tested features with the podcast and radio community in the Bay Area. In addition to usability testing, we occasionally leveraged the community's specialized knowledge. For example, while building the Audiosear.ch podcast API from the ground up, our team wanted to know: what are the most meaningful categories for audio stories?
To understand this, we brought in radio folks for a card-sort jam, using both used open and closed card-sorting to determine the top-level categories for podcasts to see how if at all they differed from the default iTunes categories. We printed out podcast covers and descriptions and asked participants to sort them into these categories, or make new ones when necessary. Indeed, we found several emergent categories over multiple users, like "Storytelling," which weren't represented in iTunes.
For show search, I designed mockups of mobile screens, and a desktop filtering flow which were passed off to an engineer, who I worked with to implement them.
After surveying listeners about podcast discovery, we heard one common complaint from listeners was that when it came to trying out new shows, it was hard to know where to start. While best of lists and curated recommendation sources abound, it takes time and work to wade through the picks and answer one simple question: what podcast should I listen to right now? PodLikeThat is a tool built using the Audiosear.ch API's "related" endpoint to answer that question.