Two Games into the 2018 NBA Finals

A Streaming TV Audience Analysis


A Quick Look Back to Set the Stage

In the 2017 NBA Finals, King James fell short, much to the chagrin of Cleveland and to the joy of the Bay Area. As the historic fourth consecutive year matchup kicked off, Conviva looked back at last year’s city-to-city matchup, not on the court but in the streaming living room which city is the King of Streaming?
2017 NBA Finals Conviva Data
The above graphic shows both the average minutes streamed per unique device or application, as well as the percentage of internet connected households that streamed each game. Unlike the box scores, each city scored wins consistently across all games. The Golden State swept the series when it came to the number of households that streamed the NBA Finals, but the Greater Cleveland Area dominated when it came to viewing time per household.


The Highly Anticipated Game One Results

While the first game of 2018 was a cliff hanger, the streaming TV results were no surprise. With over five different ways to stream the games live this year, either direct from the broadcaster’s application or through vMVPDs (such as Hulu, SlingTV, and more), and the continued flattoslightlydown audience results seen in traditional broadcast and cable options, we expected to see historic highs over the internet. Here are the 2018 Game One takeaways compared to 2017 on traffic, audience, and engagement (data is from both original broadcaster streaming as well as vMVPDs):
    • Total streams: Up 55%
    • Peak traffic: Up 127%
    • Unique Devices/Applications Used to Stream: Up 57%
    • Minutes per Unique: Up 65%
 
Net being, live sports streaming is experiencing a major spike in both number of households as well as time spent! We see this happening not only across the original broadcaster(s), but also see a boost anywhere from a 30-50% due to live viewing via vMVPDs.
 
Looking at Quality of Experience (QoE) when compared to last year 2017:
    • Bitrate for original broadcasters improved to over 6Mbps while vMVPDs saw 4-5Mbps — this could be due to more mobile / smaller screen viewing.
    • Rebuffering improved, showing a drop from anywhere between 50-75% (depending on provider) – this is a huge improvement in streaming quality and a major factor in the increased viewing time seen above!
    • Video start time remained roughly the same, ranging from 3-5 seconds spent waiting before the viewer saw the first frame of the video.
 
As far as the tale of the two cities goesGolden State’s Bay Area streaming household population increased by almost 50% versus last year’s game one. Cleveland’s equivalent populations increased by a whopping 206%. While the Cavs still trail behind the Warriors, they are gaining an internet streaming audience share 4X faster!
 

What we Learned After Game Two

While game two provided some historic oncourt performances, it did lack the cliffhanger push into overtime with Golden State dominating most of the game, but anything is possible in the NBA, especially with these two teams. As expected with a game two, some of the extreme growth numbers yearover-year slowed down gametogame (as seen in the table below citing Conviva data), but the overall total viewership continues to see significant year-on-year growth across both games. Of particular note is the fact that peak viewership across both games maintained a 72%+ growth rate year-over-year, which puts significant strain on the delivery infrastructure of streaming video publishers.
Cleveland continues to see tripledigit streaming growth and is outpacing the Bay Area streaming growth by 4X. Viewing time per unique device/application in both cities are staying inline with each other and leaving behind last year’s engagement numbers.
 

With so many options for fans to watch these games live and with the evergrowing popularity of the NBA, we expect to see overall viewership up — even though there might be ups and downs across the various viewing options. The best approach is to not focus on any one distribution channel, but rather to measure them all.