Your browser is not supported. Please update it.

5 June 2014

Video to become key differentiator for cellcos, in 15 – 20 year fight

Here at Faultline there has been something on our mind for the past few months, something that needs saying, and as usual, it is about video.

Video was the subject originally chosen for Faultline because when a digital network can cope with video, it is always a major inflexion point. In effect Faultline is about video, but at the same time, that makes it about the disruption that comes to a network when it is forced to cope with video or it finds that video is being sent over it.

It is a general rule that when a digital network is capable of carrying video, the other data on it is a small amount of the workload and there is a disjoint between pricing for data workloads, and pricing for service. Voice is a tiny amount of data for an LTE network to carry, but the utility of portable voice remains fairly high. But in order to protect voice, LTE networks have to protect themselves from being used solely for video.

It is just as inevitable that once data CAN be sent, and once network speeds are rapid enough to support video at quality, then video will become most of the traffic.

For around two years now, ever since Cisco’s VNI singled it out, the idea that video is going to swamp cellular networks has been ever present in publications, blogs and discussions. We would contend that this has always been bloody obvious.

First let’s lay out a little financial data. Revenues in Europe are down across the board, and European mobile companies are making less money. That’s not true in the US yet, but the US market still has less mobile saturation, and the competition in that market is not quite so hot as US Citizens would like to think it is. AT&T and Verizon have a virtually free ride.

Vodafone revenues are down 4.3% organically, it is projecting lower cashflow generation, European revenues fell 9%, and yet Vodafone will spend £19 billion over two years on renewing its network, during which time its cashflow after capex will be barely positive. You could ONLY make that kind of decision after someone has just paid you $130 billion, because it’s enough to make shareholders run for the hills.

Last year Vodafone added 29 million subs and in the last quarter over 13 million to reach 433 million global subscribers. All the growth was outside of Europe.

At Telefonica mobile revenue was flat, but Europe is actually down 9%, and all growth was confined to its Latin American markets, again, these are not yet mobile saturated, whereas Europe is. Margin at Telefonica was flat, profit was flat, capex was up, and mobile subscribers rose 3% to 255 million.

Deutsche Telekom owned T-Mobile bucked the trend, with revenue up 3.4%, in a market that lost subscribers. It’s capex was way up by 10% over last year, but even so free cashflow was down a huge 26%. During the year it added 12 million mobile customers, bucking the regional trends, but it raised more debt.

Orange in France has a similar story, with a 3% revenue fall, and in its native France this was 5%. And yet it now has 182 million mobile customers up 10 million. Profits fell, costs were cut, it laid off 6,000 employees, capex rose marginally to build out more LTE.

Across Europe there are more subscribers, wanting more data, and they are paying less.

The picture in the US is quite different. This is a market which has come from 80% mobile penetration a few years ago, close to saturation and customers remain enthusiastic. AT&T gained 10 million subs last year, revenue is up $3 billion on mobile in this quarter alone, while operating income was up $1.2 billion. It’s true that its net income was flat, but its cashflow went up almost 10%. Verizon is a similar story, with revenues up 4.8% and mobile revenues up 6.9%. Verizon net income was up a huge 23%, but despite this, cash generation was flat, as it added 4.4 million subs over the year,

To sum up, in a year when Cisco’s VNI says that mobile data traffic grew 81%, and video made up 53% of that traffic, subs were going up, but revenues per sub are not going up with perhaps some exceptions in the richest country in the world.

In fact this breaks out differently in each region ‘ in the US mobile data traffic grew 75%, in Western Europe it was just 57%. In the US the average mobile-connected device generated 1 GB of mobile traffic per month, up 59% from 631 megabytes a year earlier. In Western Europe it rose from 477 MB of data per month from 308 MB in 2012. Here’s how we interpret these numbers. In the US about 40% to 50% of all smartphone LTE subscribers are using their entire data caps, much of it on watching video, while in Western Europe a far lower percentage of all you can eat smartphone customers are using their entire data caps this way. This is partly because in Europe now the caps are more generous.

How can we understand this? In the US, the operator can get a fair amount of payment for its data and so invests heavily as a percentage of revenue in capex and building out more and more LTE, both for reach and capacity. In Europe one or two of the operators are spending huge amounts of capex trying to keep up with the demand for video, mimicking their US cousins, while others are taking a more cautious approach.

Right now the US operators are charging way more for data than anyone in Europe, but it does change virtually every month. AT&T has plans that go all the way up to 50GB a month, in its Share plans, but that costs $375 a month. In Germany top end 4G smartphone contracts come at a reasonable price with 20GB of data and in the UK most of the operators offer 20 to 50 GB of data, although the highest we can find on Vodafone was for 13 GB. All these have gone up considerably and that’s happened to stimulate people to take LTE devices. 3G data offerings are far lower.

We got a little bit of a clue why this week when Ericsson told journalists this week that the US was way ahead in LTE uptake and that Europe would remain behind for some years.

Patrik Cerwall, Ericsson’s head of strategic marketing and intelligence, told his audience LTE uptake in North America will account for 85% of subscriptions by 2019, but will make up just 30% in Europe despite coverage reaching over 80% of Europe.

One idea put forward for this is that the 3G networks are better in Europe and there may be something in that, but WiFi availability as a percentage of urban area is higher in Europe too. We think there is a very serious user focus on using WiFi for video in Europe right now and more people have access to WiFi for a greater percentage of their day in Europe, than in the US. So why upgrade to LTE?

But we agree that this is complex. Issues that affect it include poor data QoE, data roaming bill shock, the expensive dongle rollout in 3G, a lack of publicly available OTT video services, except until quite recently in Europe, and the reluctance for European operators to invest early enough and hard enough, in the LTE cycle, mostly because they were stung with write downs after paying too much for 3G spectrum.

In the US operators wanted to get off CDMA onto LTE, because in the long run it will bring down costs. The prediction was made as part of the latest Ericsson’s Mobility Report, which showed that LTE subscriptions worldwide increased by 35 million in the first quarter to 240 million.

The interesting thing is that none of these European operators have their own video service up and running. Vodafone has acquired Kabel Deutschland and Ono, cable companies in Germany and Spain and so at some point, you can imagine that it will offer bundled video services to mobile devices. It is also experimenting in Germany with LTE Broadcast. T-Mobile is part of an operator in Deutsche Telekom that has made a great success out of video across about 8 European countries and is approaching 6 million pay TV customers, so has much the same basis to make a similar move. Orange already has both IPTV and Digital Terrestrial TV and a huge number of unicast TV channels on Mobile, while Telefonica is in the middle of building out its Global Video Platform which targets each type of device equally.

AT&T for its part plans to buy DirecTV, and we think this deal will go ahead, even if the Comcast TWC merger is blocked, as we feel it must be. Verizon is openly talking about its plans for OTT video to portable devices, inside and outside of the home, using LTE and LTE Broadcast, and of course consumers will also take that over WiFi.

It makes perfect sense for operators, when faced with the prospect of video from companies like YouTube, Facebook’s What’s App , SnapChat and others, using up all their network capacity, to begin to offer their own video services, and we can see that this is a natural consequence of the converged pay TV, fixed broadband, mobile broadband operator, that is currently coming into vogue.

This creates a number of waves. Firstly it endorses the use of LTE for video, secondly it will set up a second round of net neutrality wars, since most of these operators will say that its own video service will not count against the consumer data caps, and thirdly it will bring a little revenue to the operators, but eat up their capacity that much faster ‘ though it may slow down churn.

Sports rights will then be bought and sold, or the companies that hold them will be acquired, just to drag more people to a particular operator’s video channel, which will be primarily aimed at driving up reliance on video data, bizarrely something that will necessitate LTE Broadcast to play at least a small part, the live event part, in video delivery.

Finally, more and more DPI efforts will be put into place to slow ABR video to something sensible, creating two tiers of video service quality.

In essence video will become THE competitive differentiator for cellular services. And the moral is ‘that if you are going to use up all my data watching video, I might as well own the service, and get some money out of it.’

Now for any cellco that has access to its own WiFi, it can use EAP SIM to offload some of this data delivery onto WiFi and alleviate much of its capex. Installed WiFi is not run anywhere near to capacity yet. Even as in the US the average home goes beyond owning 12 or 15 WiFi connected devices, so too is its WiFi technology shifting to take in high levels of MIMO, and eventually MU-MIMO to multitask sending to smaller devices.

In Europe the lower density of WiFi devices in the home is about 3 years behind the US, but keeping pace, and Asia we would hazard a guess that it’s getting close to the European device density in homes which have broadband, but a lower percentage have it.

The only thing for any given operator to decide is how to bill and authenticate services across LTE and WiFi , and there are multiple options. This will then create savings in capex, and keep the top line looking after itself.

But this has not yet happened. Most LTE operators are still getting consumer help with offload, as the consumer decides to avoid using up their LTE data cap unnecessarily, by logging into WiFi at home and at work. So far the MNO has not needed to automate this, but they will.

They must do this because the CFO cannot be kept in the dark about the costs involved. If you were the CFO at any of these major European Cellcos, then your patience with your engineering team would be running thin right now and you would be ready to overrule them or if it comes to it, fire them, for not embracing WiFi.

So while most operators in Europe say they will NOT embrace automated and seamless WiFi offload rapidly, their finance guys are saying, ‘We cannot keep up spending more and getting back less.’

And we have seen that if one operator adopts an aggressive price in a territory, offering perhaps WiFi First services, then everyone is forced to follow suit within a single 12 month cycle, or lose all their customers.

We now see that in Germany Telefonica is in talks to sell capacity to smaller German mobile providers including Unitymedia KabelBW in order to win approval for its takeover of E-Plus from KPN.

That almost obliges the weaker 3rd placed MNO, once acquired, to promise the regulator that it will support a first WiFi First service in Germany, probably with Liberty Global’s subsidiary there, Unitymedia KabelBW. The same is going on across Europe right now and launches have been effected recently in the Netherlands and Belgium, all copying France. Already the cellular revenues levels in the Netherlands have fallen dramatically.

This deal is a cinch. Liberty Global has about 50% of the cable overage in Germany, which in itself is around 50% of the market. Vodafone has the other 50%. If they partner with WiFi Homespot access, then both have 100% of the cable reach, about 50% of the country and a lot more when you calculate by population. That gives them a fighting chance of offering just as much WiFi, of a better quality, than Deutsche Telekom. The only way that Telefonica can have the same reach in Germany is to partner with Liberty Global. Telefonica has a couple of million ADSL lines with very little WiFi under its own control otherwise. Liberty Global is beginning to see that its partner elsewhere in Europe, in Vodafone, sees little reason to support it in Germany, with an MVNO in case it becomes too much of a threat.

And in the US this week, interestingly T-Mobile US says it will support WiFi calling on Apple’s iPhone, using the WiFi calling feature Apple announced in its WWDC in iOS 8. So expect more and more WiFi First services to emerge in the US around this announcement.

Any operator that cannot use WiFi to offload video consumption, has to eat up its LTE capacity faster and write off its capex investment faster, driving lower profits and a fall in value.

There are other tools for mitigation. You can build a more expensive small cell network to support video, and rely on consumer offloading to WiFi, but capex will be around 30% higher for the same result. You might invest in LTE Broadcast or in putting DVB-T2 Lite chips into phones, and have broadcast TV or in the US in ATSC/MH, although we’ve seen precious little movement in either technology to date.

While this is an interesting problem for MNOs, there is another, far more important point. If people are using more and more services for video, and that will result in making networks more and more ready for video and cheaper for data, then this in turn will drive more and more video usage and even newer video services will emerge.

When the cable industry came along in the 1960s it took 20 years before its video offering was truly competitive with that supplied by broadcast TV. Today, another 30 years on it is far better.

We have noticed that when history repeats itself, it doesn’t take nearly so long for it to happen, because those triggering it, know what they are after. And cable companies in the US were less well capitalized than cellular is today.

So we see a gradual improvement in video services on cellular. This will partly come from pay TV services first going multiscreen and then going out of home, and partly from new entrants like Netflix (everyone this week is saying Netflix won’t do so well in Germany but they provide no clear reason why). It will also come through cellular partnerships with video services.

Last week we coined the new term of Mobile First, meaning that video will slowly find that it is designed to go to mobile prior to any other delivery mechanism. That means that a slow, perhaps 10 or 15 year, decline will set in which sees professional video moving from TV First to Mobile First. We know that everyone loves TV, but everyone loved the drive in Movie, but it still died.

There will be a gradual drift to cellular operators becoming converged operators and then owning and funding both live and original content, and using that as a key differentiator to gain or retain subscribers. This drift will include further consolidation between the sectors, with operators buying pay TV operators and eventually content businesses or sports franchises or all three.

But while most people are focusing on consolidation through mergers, we would focus on the single large home screen slowly going into decline as more and more video is watched on smaller, more portable devices. Habits will change from watching together to watching separately, just as music went that same route. And the identity of the content viewer will live in the portable device rather than be attached to a home.

There will be many steps on the way. One of them is clearly going to be every single major sporting occasion and this year’s Soccer World Cup will break all records in terms of viewing figures and portable viewing figures, with almost 50% of viewers agreeing that watching on multiple devices will be really important, according to research done by set top vendor Arris which came out in this week’s Arris Video Leadership Forum in Barcelona

The company said that it expected 310 million people to tune into this summer’s football tournament globally compared to 188 million last time, with 63% watching some of it on computers, 23% on smartphones, and 25% on tablets.