Written By
John Davison, CTO
StartupLandia Case Studies

Case Study for Featured Customer

Executive Stats:

  • Consulted a non technical team and created a valuable software solution
  • Designed and implemented a custom real time voting solution

Twitch's marketing team approached StartupLandia with two problems, they needed a voting engine that could be used on live Twitch feeds and they did not have access to any engineering resources inside the Twitch organization. The essential need was the ability to ask a group of people, who were watching a Twitch feed, to vote on something using #hashtag responses. An example would be, "Which flavor of ice cream do you prefer?" Answers could be #vanilla #chocolate #mint #whatever and as people in the Twitch chat answer, an engine is counting as if "Person 1 Hey look at my phaser #vanilla #vanilla, Person 2, Oh yeah, great #chocolate". In this case, the software would count votes for Vanilla (2), Chocolate (1) and then display the results in a web page that updates live. This live updating web page could then be used as a backdrop for another live newscast like feed, something you might see on ESPN or CNN where live statistics are overlaid onto a feed of athletes playing in real time.

Something you might see on ESPN or CNN where live statistics are overlaid onto a feed of athletes playing in real time.

StartupLandia engaged the conversation by advocating for a Ruby on Rails server that could handle many tasks. In this case, it would be important to develop a web server that could listen to chat streams. This was accomplished by use of the Ruby Cinch gem, a tool for listening to irc chat servers. Second, once the server was capable of listening to a designated chat server, the server would need to be able to count votes, which also meant that the server would need to actually know what was being counted. The third major item to build here was an interface controller, something that would allow someone at Twitch to decide what voting phrases to keep track of and to have an interface that could display updated data in real time. The Twitch team also asked for the ability to toggle between different questions formats and shapes.

StartupLandia designed and implemented a custom interface that would in real time, satisfy all of Twitch's needs. The final product built gave the team the ability to control voting questions, voting options and have a dynamic interface that allowed for multiple display styles, including color and text transition.

Twitch is one the larger video traffic generators on the internet. In 2014, Twitch accounted for 40% of the live streaming traffic in the United States and 1.8% of all internet traffic, second only to Google, Netflix and Apple. In 2017, it surpassed legacy network ESPN in audience size and live streamed more content than ESPN, WWE, and ML-combined. It is difficult to quantify the total reach of a project like this, but the total possible reach is basically analogous to a major section of a large ocean.