Your browser is not supported. Please update it.

3 July 2014

Verizon boosts CDN, but cellular letting video revolution slip away

Verizon has updated the world on its progress in improving the capacity of its Content Delivery Network (CDN). Using its recently purchased EdgeCast assets, the company is investing in safeguarding its service against the inevitable future growth in video data consumption. It is not too speculative to think that it may be taking the initial steps towards launching a global OTT video service, if it can secure attractive content deals.

Verizon said it has been expanding its CDN capacity gradually since January, by adding 20 new points of presence (POPs) in major cities including Batam, Helsinki, Jakarta, Kaohsiung, Melbourne, Milan, Sao Paulo, Stockholm, Vienna and Warsaw. Extra POPs have been added in cities where Verizon already has installed CDN assets, including Amsterdam, London, Madrid and Paris.

A CDN works by caching a copy of a file closer to the edge of the network and the end user rather than storing and accessing a single copy at the heart of a network. This network architecture reduces the bandwidth constraints on the central server and will help alleviate bottlenecks in the distribution path, improving the viewer’s experience.

In addition to EdgeCast, Verizon also purchased Intel’s OnCue project after the silicon giant gave up on plans to launch a cloud TV service. Verizon has also bought the uploading and encoding services of upLynk outright, ‘whose exclusive technology streamlines the process of uploading and encoding video for live, linear and VoD content.’

If you put all the pieces together, you’ll see that Verizon is poised to launch a rather revolutionary TV service ‘ entirely on the back of its 103 million mobile subscribers in the US, but it appears that this service could now be delivered globally. Verizon’s network investments mean it is perfectly able to launch a nationwide video service, not restricted by the geographic limits of its FiOS lines. It would be a serious blow to its rival cable competitors who are starting to see cord cutting to internet-based video services like Netflix.

We suspect that Verizon will combine its existing content deals with any that the Intel purchase brought with it, alongside the rights it has acquired to the NFL, as well as the content available from its Redbox Instant video service, to package and sell to its mobile customers and FiOS homes.

But what’s key to this strategy is fundamentally (and gradually) changing the way that TV subscriptions are bought and sold. Traditionally, a home is provided with a satellite dish, or a cable or fiber connection and TV is piped in for any occupant to consume ‘ limited only by the number of set tops and TVs. In the last few years, TV operators have been providing TV Everywhere services to customers, allowing them to access the TV content they pay for outside of the home ‘ a service which many viewers feel entitled to, given that they have paid for content, but don’t want to be bound to the TV to watch it.

Central to this aim is the smartphone, both as a viewing medium and more importantly as a means to identify a user and a valid subscription. The SIM card in a phone solves the issue of authorization and security and allows a user a truly portable TV experience.

If operators get on board, and this may be the largest stumbling block for Verizon, a user could ‘cast’ content from their phone to any screen they are sat in front of. Inside the home, this fuels increased consumption by letting a viewer watch on the screen of their choosing, but when away from home the viewer can share pay TV content with their friends ‘ something that will undoubtedly initially scare rights holders.

It might take quite a lot of persuading, but they may be convinced that any time the viewer is away from home and not watching TV, valuable revenue opportunities are slipping away. If the viewer is given the option to watch any TV content in any location and at any time from a single service, it seems a logical conclusion that viewing figures would increase and subscription revenues as well as ad spots could become more lucrative. As both are inextricable from each other, TV operators and rights holders should be working hard to let their customers watch as much video as possible ‘ regardless of location.

And in that final point, Verizon’s CDN investments and cellular network appear to place it on the verge of realizing this service, riding the crest of a wave in the new millennial mobile-centric generation that will help them try to wrest video ownership from cable and satellite ‘ until you realize that between them, AT&T and Verizon have less than 50,000 WiFi hotspots and so delivering that video could get tricky. Let’s put that into perspective.

The new LTE cellular networks of AT&T and Verizon are capable of supporting video if they can guarantee that their customers are using a phone with enough radios in it to support the fancy OFDM and carrier aggregation techniques required to push lots of video to lots of handsets simultaneously. But that phone costs a lot to make, thanks to the extra radios, and of course the cellular operators also have to invest in a new, increasingly dense LTE network.

We have been pushing the idea that WiFi is their best chance at what we are calling ‘mobile-first video’ or mobile centric video services, but both AT&T and Verizon are trailing woefully behind Comcast, which has been steadily accelerating the pace at which it upgrades home routers to WiFi gateways with both a public and private SSID. For every house Comcast owns, it potentially has a homespot with which to power its wireless network ‘ and Comcast has already switched on 3 million of them compared to the cellcos’ 50,000, and will add another 5 million this year. If anyone is empowered to build a mobile centric video service which relies on WiFi, it is Comcast, while AT&T and Verizon are forced to continue the expensive build out of LTE and LTE Advanced.

When you add the 250,000 or so public WiFi hotspots brought to the Comcast table through the Cable WiFi Alliance, (remember hotspots are deployed to maximize coverage unlike the vaguely haphazard deployment patterns of homespots ) it becomes apparent that Comcast is best placed to push a wireless mobile-first network ‘ a position bolstered by its strong content agreements. It could launch video on the move to virtually anywhere, and not even build a single cellular base station.

And this means that Comcast is best placed to take the mobile-first approach to viewing ‘ ironically on devices usually provided to consumers by Verizon and AT&T. That’s how poorly the two largest MNOs have forecast the demand for video. They hoped to do it all on their own network and are being undercut by WiFi coverage. To build out 8 million small cell base stations is going to cost them considerably more than those WiFi homes that customers have paid Comcast to build out.

But there are other factors at play here. The cellular operators want their customers to keep paying for data. In this light, they want their customers to access video over their network, not from WiFi and not from files downloaded to the device over WiFi. One simple solution to the problem of roaming and intermittent connectivity is to download the video files onto the device and protect it to a standard that appeases the rights holders. This allows a user to seamlessly consume video on the move without the MNO worrying about a negative customer experience when things go wrong and the connection hangs or stutters.

This would require storage capacity in smartphones to increase significantly, which may seem unlikely, given the incredibly slow rate of improvement seen in the flagship phones from the likes of Samsung and Apple, and of course the MNOs won’t be clamoring and hounding OEMs for larger storage as the MNOs still have to pay for those phone and anyway want the consumer to pay for cellular data and not pre-load their entire day or week’s viewing onto the internal storage of the phone.

So it looks like phones won’t become portable DVRs anytime soon, unless SD card support gets simpler for consumers ‘ and Apple shows no sign of beginning to support removable storage.

We feel that cellular operators are wasting their opportunity to seize control of the TV market from cable operators. No amount of CDN investment makes up for the issues that cellular delivery of video content at scale will suffer in its current deployment. In the US at least, it looks like we’ll have to wait for Comcast to launch the first truly mobile-first service.