- The worldwide market for “Over the Top” services was estimated to be $ 28.04 Billion in 2015 and is slated to reach $ 62.03 Billion in 2020 according to the research firm, Markets and Markets.
- eMarketer reported that in 2015, 181 million people in the US would have watched streaming video content over the internet on their phones and that this number could all-but-touch 200 million by 2019.
- A study in mid-2015, part-sponsored by Ooyala, projected that the USA revenue from Premium OTT Video would range between $ 8 billion to $ 12 Billion on 2018, with NetFlix being the largest player.
- In Asia Pacific alone, Media Partners Asia, recently forecast that paying subscribers of video-on-demand services would rise from 177 million currently to over 360 million by 2021 with some estimates of revenue being as high as $ 35 Billion in Asia alone.
Any way you slice it, these are substantial numbers. The significant, and sudden, improvement in the usable bandwidth and enhanced visual capabilities of mobile devices are contributing to this push to consume video-on-demand but what is really driving this movement is content. YouTube and NetFlix rule the roost but new players with specific thrusts in programming are springing up every day in newer and newer markets and making an impact. For instance, in India, more people had watched the English Premier League on the app Hotstar, than had watched it on TV screens. OTT Video is here to stay – that being the case, is testing the end to end solution that delivers this rich video experience different? It seems clear that how video is delivered over OTT is significantly different from how streaming video was delivered over IP networks. Now video is delivered as small files rather than as a bunch of packets streamed. The biggest difference is that the videos are consumed, and hence delivered continuously and this suggests some specific considerations that should be addressed while doing this testing. Let’s look at some of these key considerations.
- An integration of systems: There are several players involved in this game. This starts right from the content originator, who needs to maintain an appropriate level of encoding quality at the origin. Then there are multiple HTTP servers that come into the picture, and how these are available to the service and their performance all impact the end quality of the video. Another significant element is the network – how good and fast it is, how much time does it take for the video while in transit and the usable available bandwidth are all potential points of failure and hence need to be accounted for while testing. Given the impetus to monetize, ads play a major role in this system and that suggests another point of integration (failure?) to test for – the ad server. Another piece of the puzzle is the user. One obvious element to test for is the user client – how it addresses issues like buffer management, time between video elements and so on. The devices are a big part of this story, but more about them later. That apart there are the issues related to user management to check for like securing the users access, logging of the specific amount and kind of services the user consumes and billing for that are generally within the purview of billing servers – another key integration point to test. While testing each element individually is important, testing how they work together to deliver a seamless, continuous experience is even more critical.
- Devices: The market is awash with smartphones and tablets with innumerable form-factors and software capabilities. Then there are the various game consoles, smart TVs and other players and devices that can be used for consuming OTT Video. Given the differing capabilities of each of these and the specific behavior of each device, it becomes important to test the video service across multiple devices. Physical availability of devices may become a constraint, which is where simulators can play an important role.
- Quality of Experience (QoE): This has become more challenging in the days of larger video files. Earlier measures of the video QoE were designed for measuring extremely short videos. Attempts are continuing to define standards that can be applied to the longer duration videos delivered over OTT Video and these are largely driven by the Video Quality Experts Group’s (VQEG) inputs to the ITU. Establishing the standard can then promote identifying the QoE at each stage of the lifecycle of the video and to then pinpointing where a loss of quality is taking place. Typical issues related to the QoE are the video stalling while playing, a lack of sharpness or clarity in the video, pixelation or blocking in the video, jerks in the video, lack of clarity in the audio or a lack of synchronization between the video and the audio and the complete stoppage of the video.
- Spurts in demand: This is a problem specific to the OTT Video space in many ways and is related to an explosive demand for a new piece of content as soon as it becomes available. This may be because of it’s news-worthiness or because it is highly anticipated. The challenge to those testing the OTT Video system over which such content will get consumed is how to simulate such an extreme spurt in demand? It’s not only a question of testing for the performance of the servers but also to test for what happens in instances of curtailed bandwidth and a large number of viewers logging in simultaneously.
OTT Video is already seeing an explosion of interest and as higher bandwidth networks like 4 and 5G go live around the world chances are the on-demand video consumption will only go even higher. To those responsible for ensuring that the end consumer gets a high-quality video experience – maybe this post will help you draw up your list of testing priorities towards that goal?