How Pluto TV Pivoted From AVOD to FAST
Learn more about AVOD and FAST at Streaming Media East.
Read the complete transcript of this clip:
Eric Schumacher-Rasmussen: When you decided to make the pivot from AVOD to FAST, what technological changes did you need to implement? And could you talk a little bit about the decisioning process? I mean, by the time you made that pivot, you had a little bit more money, but I assume you're still very much in that that startup mode, even though you've been around a number of years. Talk about the decisioning process, talk about the technologies that you explored and how you ultimately decided what you built yourself versus what you perhaps turned to third parties for.
Vibol Hou: Right. So, we go back to customers, and in our formative years, our ratings were not great. Two and a half stars, three stars. It was pretty bad. And the feedback was, it keeps buffering. So we're doing client-side stitching, which means that every clip that gets loaded has to buffer, even if it's a fraction of a second, there's still a pause in between the video, whether it's the video itself, we also have to integrate with programmatic ad technologies that were built for pre-rolls. The mid-roll ad insertion on the client side was just not that great. Also, wait a minute and see the spinner, wait for that ad to come back from the networks, and then play it back.
And that was just a very clunky experience and we thought to ourselves, we looked at technologies, OVPs on the market at time, and some of them were capable, but they were still outside our price range. They were demanding per-channel fees and platform fees that we still couldn't afford. Literally, we weren't able to put the money down to partake in the services. And so we decided to build it ourselves. So we started to figure out, how do we compose these virtual live linear streaming channels? We took a look at all of the specifications, the HD live streaming spec. Everyone's pretty familiar with HLS. And we started to concept out stitching VOD assets together on the server side with those specs.
Once we proved that we could do that, we knew we were on to something. We were able to start expanding on that principle. We built our ad proxy ad insertion systems with the Video Stitcher. We had our CMS and all the pieces that we needed, and then we needed to upgrade our clients. Having all that code, doing the client-side stitching made the clients quite a lot bulkier. And we tried to normalize on web technologies. But again, at the time these web views were inconsistent, particularly on the Android platform. There are too many generations of web views, and we couldn't support all of them. We tried, and we built our own chromium layer on top of the Android app, just to be able to try to render all of the elements of video that we needed to render. And it took a lot of time, but once we got to the server-side stitching and everything was a single stream and the video player to load that and play it back seamlessly, there was no pausing in between assets. You just had a brief moment of spin up on the front end, and everything else was just streaming. There was adaptive bit rate, all the things that we needed to improve our experience. We were able to build it in just about six months.
This potential for innovation and accessibility truly sets FAST platforms apart from both traditional over-the-air linear and on-demand streaming platforms. Over the coming years, we will increasingly see that impact as more device manufacturers and VOD platforms, both ad-supported and paid, enter the space, providing an ever more diverse set of curated and personalized linear channels to their users.