The 2014 FIFA World Cup is breaking all records for live streaming, according to data from Akamai. This week’s first-round match between Germany and Portugal drove a peak of 4.3 terabits per second over Akamai’s content-delivery network, easily eclipsing the previous record of 3.5 Tbs for the U.S.-Canada men’s hockey semifinal during the 2014 Sochi Olympics. The USA’s opening match against Ghana drove 3.2 Tbs, just behind the 3.4 Tbs for the Spain-Netherlands fixture.
In terms of actual eyeballs, Germany-Portugal attracted 1.1 million online views for ESPN’s WatchESPN app while Univision saw 2.0 million digital viewers tune in. US-Ghana drew 1.4 million online viewers on ESPN and 1.7 million for Univision. Millions more have been streaming the games outside the U.S.
The data underscore the growing demand for live-streamed events, especially live sports. As I discuss in an upcoming report for GigaOM Research, that demand is being fueled by a variety of factors, from the increased use of mobile devices to watch video, especially tablets, to the deployment of new technical standards and technologies by service providers (including Akamai) to better enable live-streaming at scale.
The tilt toward live events on over-the-top platforms is also being driven by many of the same factors behind a similar shift on traditional pay-TV platforms. In an age of highly fragmented audiences, live programming commands far higher levels of viewer engagement (and far lower levels of commercial avoidance) than time-shifted and on-demand content. According to Ooyala’s 2014 Q1 Global Video Index report released this week, viewers watch live content for nearly an hour per play on connected TVs compared to only 4.6 minutes on average for on-demand content. On desktops, its 34 minutes for live compared to 2.6 minutes for on-demand, and on tablets the ratio is 9.8 minutes to 3.7.
Spiraling rights fees, especially for sports, are also putting a premium on exploiting those rights on as many platforms as possible.
Perhaps the biggest factor likely to continue fueling the growth of live-event streaming, however, is the emergence of better monetization tools. Dynamic ad insertion, coupled with the ability to insert unscheduled ad cues into live streams, are enabling OTT providers to strip out the broadcast ads and replace them with targeted, measurable ads in real time, creating a new revenue stream against the same content.
According to a study released this week by Magna Global, the research arm of ad-giant Interpublic Group, digital media advertising revenue grew 17 percent in 2013, to $43 billion, and now surpasses national TV in its share of the total advertising pie. While search and social are still the biggest chunks of digital advertising, mobile and video are the fastest growing pieces and responsible for nearly all of that 17 percent increase.
Those numbers have clearly attracted the attention of Google, which has lately made a series of moves clearly aimed at positioning itself to dominate live online video advertising just as it does with search advertising. Last month, reports surfaced that YouTube was in talks with Twitch.tv about acquiring the live gaming site for $1 billion. So far, at least, nothing seems to have come of those talks, but assuming they were real at one point the deal would represent a major thrust by YouTube into the live-streaming arena. According to Qwilt Analytics, Twitch accounted for nearly 44 percent of all live-streaming traffic by volume and 1.8 percent of all downstream internet traffic during peak hours.
Then this week, Google announced the acquisition of ad tech firm mDialog, which specializes in dynamic ad insertion in live streams. In a blog post, Google said it plans to incorporate mDialog’s “technology and expertise” into its ad-placement service, DoubleClick. In a post on its website, mDialog said, “we’re looking forward to offering content creators new and even better ways to make money from their live and on-demand content.”
Google’s interest in Twitch and mDialog comes on the heels of a series of deals, unveiled at the International CES in January, with a broad range of TV manufacturers and chipset makers to embed Google’s VP-9 codec in their 4K Ultra HD TV sets and system chips. Like the HEVC codec developed by the Moving Pictures Expert Group standards-setting body, know officially called H.265, VP-9 is being touted as a compression format for 4K video. And to attract interest from TV makers hungry for 4K content to drive sales of new Ultra HD displays, Google announced at CES that it has begun to convert the vast YouTube catalog to VP-9 and will begin streaming content using the codec in 4K.
As I discussed in a blog post at the time, however, Google is merely using 4K as a Trojan. While both VP-9 and H.265 can be used for 4K they can also be used for regular HD or SD video, with a roughly 50 percent bandwidth savings compared with the current H.264 standard. Given the enormous amounts of bandwidth required to stream live video at scale, that kind of bandwidth savings is critical.
Unlike H.265, moreover, which comes with hefty licensing fees, Google is giving away VP-9 for free, with few licensing restrictions, which makes it a much easier lift for device makers. While some have speculated that Google’s support of its own next-generation compression system could set off a format war with H.265, I think Google is far more interested simply in getting a high efficiency codec to support live streaming deployed as quickly as possible. The high royalty stack associated with H.265 is bound to slow its rollout, whereas the royalty-free VP-9 can be deployed today, both at the device level and at the encoding level, without having to worry about cost (if and when the time comes I have no doubt YouTube will also support H.265).
Connect the dots and what you get is Google putting in place the pieces of an over-the-top, live-streaming ecosystem, including device support, content monetization, an optimized streaming platform and, if it were to acquire Twitch, an audience of 45 million unique monthly viewers.
If I were a traditional pay-TV provider or linear broadcaster, that would worry me a lot more than Netflix or Hulu do.