Creating A Conversation For Video


Last October, we released what I see as one of the more impactful pieces of video technology for a digital publisher in recent memory. This tool is about conversation and context.

A year or so back, Ayalla Barazany and I were discussing video, our audience & how to increase overall engagement. We were well aware that the content was the real key to engagement but from the technology side you don’t have much say in how it’s produced. So we set our eyes on the player, the experience & our users, all things that we could effect. Ayalla was previously at Travora Networks and mentioned an idea that she had related to annotated travel video. The conversation rapidly transformed into time based comments with in video. We knew from YouTube that people were already talking about specific moments in the comment sections. Action sports video is where most of my viewing is focused and the comments are rarely about the holistic experience, rather an exact second where something happened.

This is a pretty standard comment thread in a skate video. Users calling out tricks, songs or anything they see at a very specific moment in the video.

We wanted to create contextual comments that were more deeply rooted into the experience. This was all just a rough concept/hallucination with no proper validation.

Not long after that conversation, I took part in Complex’s first hackathon. It was relatively freeform, with one goal, to create something and have fun. With an 8 hour period dedicated to hacking, a few eager developers and myself we were able to flesh out a proof of concept for that would act as a baseline for 2 months of development to complete REACT, a contextual commenting system.

At it’s core the product is very simple, contextual time-based comments in within a video player. Utilizing Ooyala’s Player API for a custom skin and Disqus to create a foundation for the comment system, we were able create a tool that enhances the conversation across our digital landscape.

The tech was simple, the UX was a nightmare. How do you make a conversation contextual to the video, keep it in view but not detract from the overall viewing experience. And don’t forget to solve for mobile, it’s only 75% of your audience. For weeks our lead designer & myself destroyed the whiteboards during user flow, wireframing & heated scribbling sessions in order to solve the most complex design problems. What we came out with was minimal & clean when you consider the functionality included.

The official launch of REACT was a part of the broader launch of Complex’s editorial joint venture with Tony Hawk, The RIDE Channel. Shortly after the launch we posted a video of Tony Hawk riding the Hendo Hoverboard and we got to see REACT in action.

Despite some of the initial engagement with REACT I wanted to see much more consistent usage & more engagement on each video. From my experience new ideas & technologies require a significant amount of education & time to get over the learn curve and gain broad adoption. We knew this going into the product build.

After launch I identified a few major barriers to adoption:

  • Player had a distinctive look but no introduction as to way it was different
  • No social authentication for login, Disqus API only allows for standard username/password login
  • No intuitive direction on how to post comments

The solutions:

  • Customize the video loading animation with subtle introductions to the video player
  • Integrate intuitive social authentication that removes any additional friction for users to post a comment
  • Annotate the player controls to show users how to post comments(see screenshot below)
REACT Player control annotations

Like all products, continuous iteration is required in order to make it truly successful and with REACT it was no different. Within days of releasing the product updates we began to see significantly more engagement across all videos.

Will REACT be a catalyst for similar products or will another product be introduced that will better contextualize conversations in video?