March 2012

new iPad heats up race for 700MHz in Canada

new iPad heats up race for 700MHz in Canada

The new iPad

The new iPad came with many exciting features, but for Canadian operators it also came with more issues, possibly creating a game changing tussle for 700MHz in Canada.  Yes, the resolution, processor and camera are all amazing, but buried deep in the very exciting LTE wireless part of new iPad was a good reason to fight for 700MHz spectrum.

In the USA there are two versions:

  1. Wi-Fi + 4G for AT&T model: LTE (700, 2100 MHz)3; UMTS/HSPA/HSPA+/DC-HSDPA (850, 900, 1900, 2100 MHz); GSM/EDGE (850, 900, 1800, 1900 MHz)
  2. Wi-Fi + 4G for Verizon model: LTE (700 MHz)3; CDMA EV-DO Rev. A (800,1900 MHz); UMTS/HSPA/HSPA+/DC-HSDPA (850, 900, 1900, 2100 MHz); GSM/EDGE (850, 900, 1800, 1900 MHz)

Three things really stand out:

  • There is no HSPA in AWS, so without launching LTE, the new entrants cannot have iPad subscribers
  • The Verizon model has no LTE in AWS so LTE is ONLY in upper C block of 700MHz (band 13)
  • The AT&T model has LTE in lower B + lower C 700MHz blocks (band 17) and in AWS

700MHz Implications for Canadian New Entrants

Assuming there is no set-aside, the Canadian incumbents will get Lower B, Lower C and Upper C blocks in 700MHz, so new entrants will be relegated to second class 700MHz in Lower A at best.  Even the Upper C block was split into two parts, this will be an issues for new entrants.

700MHz Implications for Incumbents – Bell, Rogers and TELUS

Since all three Canadian operators will be deploying the AT&T version of the iPad, which only has lower B and C blocks in 700MHz, three operators will be bidding for two blocks.    This means that the customers of at least one of the incumbents will disappointed when they can never use their iPad in rural areas on 700MHz.  There are winners and losers, with no middle ground, as switching to the Verizon version would be pointless as it does not have LTE in AWS, making today’s LTE investment redundant for the losing Canadian operator.

So who will win the lower B and lower C blocks of 700MHz?

Historically Rogers has always won the best spectrum.  They have paid more and almost without exception paid whatever it takes to win the best spectrum.  As a result they have always enjoyed a spectrum rich environment and share a similar ecosystem to AT&T.

But in a make or break situation, where Bell and TELUS have more to gain by winning two blocks of continuous spectrum that they could share across chipsets and networks, one has got to believe they will not give in without a fight?  Bell and Rogers have enjoyed significantly lower capital intensity (10% and 11%) recently versus Rogers at 16% even while rolling out LTE.  So they have some extra cash sitting in their back pocket.

We hope that by raising dividends, that none of the incumbents have given back too much when they might need the extra funding in what now looks like a significantly more competitive 700MHz auction, even without special treatment for the new entrants.

The race for 700MHz really has had the heat turned up, by Apple’s new iPad no less. (c) Alphasynb


No comments

iPad 3


Yesterday we read a really interesting article by Jameson Berkow iPad 3 has tablet owners wanting to upgrade in then FP TECH DESK.  The article points to a survey saying that 42% of current iPad 2 owners plan to upgrade once the device is released.  We think this is an incredible situation.  Gone are the days when subscribers are loyal to their carrier, now it is all about the devices.  This cult-like following has many implications and causes concern at many levels.

innovation adoption curve

Rogers Innovation Adoption and Diffusion Curve

Firstly, Apple products are defying the most sacred of business principles – that of Rogers, who developed the product adoption curve and diffusion of innovation (no relationship to the Canadian carrier).  Rogers did over 500 studies on innovations that all showed the progression from innovator (2.5%) to early adopters (13.5%), to early majority (34%), to late majority (34%) to laggards (16%).  This bell curve (also no relation to a Canadian carrier) has held steadfast since its publication in 1962.  But is 42% of current iPad 2 owners plan to upgrade, this theory could be completely shattered?  Product life-cycles have also been dramatically shortened.

Secondly, this changes the ability for retailers to sell the iPad 2 since the market will almost certainly be flooded by less than 1 year old iPad 2’s that will probably all be in perfect working condition.  Also retailers normally sell on price (which is fixed) and features of one product vs alternatives, but in this situation, it seems that the masses will buy it no mater what the specifications, so only availability determines a retailers profitability.  Those who have good relationships with Apple will enjoy more profit.  This is also a shift in power from retailers to manufacturer(s).

Thirdly, while US and Canadian subscribers have become accustomed to handset subsidies, including on expensive iPhones, this shows that iPad fans, who surely overlap with iPhone fans, are willing to spend their money on unsubsidized devices.  This could become another loyalty problem for carriers, who have retained customers by locking them into long-term contracts in exchange for handset subsidies.  Carriers should discount the rate plans of customers who choose to bring or buy their own handsets.  Shorter product life-cycles also put carriers, who are trying to lengthen associated contracts, churn strategies in jeopardy.  This is a shift in power away from carriers to handset manufacturers.

If the new iPads do have LTE, watch how this will turn up the heat on LTE deployment globally, shifting the balance of power to carriers who have the financial strength and spectrum to deploy quickly.  With the unlocked, unsubsidized iPad 3, LTE laggards will see all tablet users defecting to better networks.

Finally, this pre-launch hysteria is bad news for RIM.  RIM recently did the long waited and much delayed upgrade to their tablet Playbook OS 2.0, but we have not heard much buzz since the press release, which seems only to placated people who already bought the playbook.  Now playbook users can do email on their playbook!  And RIM is due to release Q4 results on March 29, but earlier this week, Jefferies Analyst Peter Misek warned that there was a greater than 50% chance RIM would negatively pre-announce 4th quarter earnings, which most analysts believe could miss expectations.  Apple’s timing seems immaculate; remember that they launched the iPhone 4S after RIM had a 3-4 day global outage, where nobody received emails.  Seems like Apple even has the luck of timing on their side.   With the iPad the apple of consumers’ eye, RIM’s tablet could be relegated to oblivion.

So Apple is not just taking extreme market share from it’s competitors, it is taking loyalty, power and revenue from carriers, revenue and power from retailers and is systematically re-construction the value chain in wireless.  We love Apple products, but would also like to have a choice.  Wireless looks like it is on a similar path to music, which was a big, competitive and long established market with significant players, but now in dominated by Apple.  Make no mistake, consumers did benefit from cheaper and more convenient music, but artists do not make more, there is less industry investment in new talent (we don’t have facts to back this up – it is just speculation), the labels are a shadow of their previous incarnation and we will leave it to better critics than ourselves to decide whether music is better or worse?   We are not sure that it is reasonable to prevent the at&t merger with T-Mobile USA while sitting on the side lines while value transfer moves from all players to one company.  Perhaps the enlarged at&t and a strong Verizon would together have been enough to at least maintain the balance of power between Apple/Google and at&t/Verizon?

The wireless industry is changing more than in many years.  Look for the once mighty carriers to be dis-intermediated and content and device players to contribute more in the overall value chain.  Good news for Apple and some consumers, maybe not so for RIM and the carriers.

No comments

Unlimited wireless plans?

Unlimited plans

From a marketing perspective it is always very attractive to have unlimited, but delivering unlimited is always difficult when the resource itself if not unlimited.

Have you ever heard of an unlimited gas plan or unlimited electricity?  We have not; in fact it looks like the only thing that we do not pay for in measured quantities is air.  And even with air, we have recently read many articles discussing how people from countries with air pollution problems are immigrating to North America for improved air.  So even air might have a levy or volume based rate at some point.

In telecoms the fixed costs are high and the marginal cost of delivering bandwidth very low, but not free.  If the capacity of your network is sufficient to handle all the traffic, then it makes sense to price additional usage very low, but this is not the case when incremental investment is required.  Then the type of service really matters as well as the source and destination of content.

For example, in circuit switched voice, true each conversation takes a whole circuit, which is expensive, especially if there are silent periods, but then at least there is a limiting factor – it takes at least two people to be present for the service to be used, and since most people need to eat, sleep and talk to other people, it does not happen that they can occupy that circuit continuously for 24 hours a day.  Note that there was a notable exception to this in the UK, when a new wireless provider gave away an extra handset with every subscription and offered unlimited (with no caveats) calling between them.  The urban legend says that many families took these plans and used the devices 24/7 as baby monitors, allowing parents to nip out to the local and still be constantly listening to their young ones.  Most operators have learned from this and unlimited plans generally have a fair use policy, which allows the service provider to disqualify obvious abuse.

But data is different in a number of ways, some good, some bad for the provider:  On the positive side is the fact that a whole circuit is not used and even VoIP calls use the network much more efficiently than their circuit switched alternative.  On the negative side, data does not require two people to be present at the same time. Indeed massive amounts of data can be moved without anyone consuming it.  One example is streaming content vs broadcast.  In a world where homes are increasingly moving away from cable and satellite towards an IP-based world of streaming, the economics for service providers can get ugly quickly.

Cable companies can choose not to provide a particular channel at a node level if nobody is watching it allowing them to re-use that bandwidth, if one home in the node is watching the channel, there is no incremental cost of the bandwidth if a second home starts watching the channel.  In streaming, the math is very different.  Each stream will use incremental bandwidth.  This is fine if they paying for the stream in a pay-per use model, but in a world where catalog content’s margin cost is zero (through Netflix for example) then incremental network is used for every viewer.  This get really nasty when all content is online, for example, when a home is streaming live TV, live radio and other content at the same time.  But when someone leaves the room, or forgets and leaves a stream on all day, then the usage can get extreme.

Data today is also more like voice in the past in one way however, it is becoming more symmetric.  So ADSL and cable DOCSIS 3.0 networks that have been built to deliver content to homes, does not have enough upstream capacity to deal with the new generation of user generated content.  How significant is this?  Very.  Over 60 hours of video are uploaded to YouTube every minute.  Over 4 billion videos are viewed every day.  There are 800 million unique visitors per month.  Finally, more video is uploaded to YouTube every month than the 3 major US networks created in 60 years!  And this is just YouTube.  So our networks do have to change to accommodate.

USA Canada Spectrum Holdings

USA Canada Spectrum Holdings

Wireless networks are much better at uplink – indeed many carriers in north America today offer faster uplink speeds over mobile than they do over ADSL or Cable.  So speeds are great, but bandwidth is very limited.  Cable companies have between 650MHz and 750MHz of bandwidth per household, Verizon has 87MHz, AT&T has 89MHz, T-Mobile has 50MHz, Sprint has 35MHz, MetroPCS 19MHz and Leap 16MHz, Rogers 101MHz, Bell 61MHz and TELUS 68MHz.   (Note that all of these are averages of major markets as any carrier might have more in one market than another. )  New entrants in the USA and Canada have significantly less.  So there is not unlimited capacity even with spectrally efficient technologies like HSPA and LTE.

Carriers in the USA and Canada launched with unlimited plans and now have three problems:

  • They cannot continue to build networks fast enough for the demand
  • There is not enough spectrum for the number of carriers
  • They have set expectations with customers that unlimited is desirable

This is further exacerbated by the fact that all carriers have a few users or ab-users that use the vast majority of the traffic.  If their usage was moderated, the rest of use could happily survive of today’s networks assuming reasonably quick allocations of spectrum and additional carrier investment.   AT&T recently suggested that the top 5% of users in their unlimited plans used 50% data than the top 5% of users in tiered plans.  Now they are resorting to throttling these users to lower their usage, but it is always difficult to go back on what you originally promised.   Unfortunately it also looks like some of the non-abusers were effected by the throttling, which has gone down like a concrete parachute.

Bandwidth is a limited resource and while carriers should not try to overcharge as a result, we should all pay in line with our usage, but at moderated unit costs.  Sure one should get volume discounts if you consistently buy a lot, but unlimited plans self-select ab-users and everyone else has to pay for it.   Let’s all work together to re-align thinking around delivering the scare resource of bandwidth in a fairly priced manner.  This means we all need to reset our expectations, abusers will have to pay more, governments will have to allocate more spectrum and carriers will have to invest more.   Finally carriers can reduce their investment by sharing parts of the infrastructure in a wholesale and retail market.  (c) Alphasynn

No comments

Cutting the cord

Cutting the cord

Recently much has been written about cutting the cord.  This phrase can refer to a number of situations, but the two most obvious in today’s parlance are either dropping your home phone or your cable TV.

At home, we have done both and it has been amazing, liberating and has had the side benefit of transforming our family off the couch and into the other rooms in our home.  The interesting thing is we spent over a decade in the UK, where only 12% of home have cable TV – in fact over 50% of home have free digital terrestrial TV.  In the UK there is an annual TV license that helps pay for this and also for creating content and everyone has to pay it, even those that have cable.  In the UK there is also more fixed-wireless substitution.  At least 12% of houses in the UK only have a mobile phone (this is a 2009 figure, but we could not find an updated number).  This is partly because it is free to receive calls on your mobile phone in the UK in a caller pays model.  This makes the barriers to being a mobile phone subscriber significantly lower and unused wired phones at home are disconnected.  But the number of home phones is also much higher than it needs to be because BT dominates the broadband provision in the UK and they require you to have a home phone to have broadband.  (Or at least the discount means you pretty much have to take it.)  Do we use the TV, absolutely, we have Netflix have two AppleTVs (being ‘early adopters’ we got them both on the first day of their launch) and there is a lot of legal content streamed on the internet.  (We do not download pirated movies or music).  We watched the Superbowl live and had the choice of 5 camera angles and got to watch the USA adverts to boot!  Increasingly content providers will go around the same expensive cable and satellite solutions that consumers want to avoid and the telecoms and cable companies can focus their attention on providing really good internet.

At the beginning of March Comscore released their 2012 Canada Digital Future in Foucs.  The Comscore data reaffirms that Canada’s population spend more time on the internet and use the internet more than people in any other country.   The skeptic in me wants to say that we spend more time on our internet because our internet in Canada has the unique characteristics of being both one of the most ubiquitous and the slowest.  So it is everywhere, but slow due to the lowest levels of fibre to the home in the OECD.   On a more positive note, nearly 100% of homes have internet that is fast enough for VoIP.  Open standards-based VoIP solutions should be used at home and on your mobile phone.

What Bell, Rogers and TELUS have done is invest in wireless.  They have all deployed LTE in most urban areas and we would surprised if we don’t have near the highest population coverage of HSPA+ anywhere in the world.  But these are data networks, not voice networks.  These high speed data networks are often faster than copper wire and cable based internet and this week, the FT reported that Telefonica demonstrated it could consistently deliver 76Mbps at the MWC in Barcelona.  This is most than 1,000 times the speeds needed to offer a good VoIP call with better fidelity than either a landline or normal mobile phone call.   So let’s use our super fast Canadian data networks for voice!

Canadian cord cutters

Canadian cord cutters

The home phone market has changed a lot over the last few years and in Q4 alone over 120,000 customer left the incumbent telecoms companies (Bell, Bell Aliant, MTS and TELUS), but did NOT take a new service from any of the cable companies.  Are these subscribers going wireless or to alternative VoIP providers?  We are not sure, but we suggest you do both.  There are VoIP solutions that allow customers to download a smartphone app and use their smartphone for good quality VoIP calls over wifi, 3G or 4G networks.  Finally something valuable that can use up your mobile data plan – voice!  © Alphasynb

No comments

Q4 2011 Wireline Results – Canada


In Q4 we had some interesting activities in Canada.  TELUS had some exceptional results, but otherwise all trends stayed the same.  Firstly wireline phone:  Overall nets for cablecos were lower than Q3 2011 but higher than Q4 2010.  The real cableco winner was Cogeco, which has 24k nets vs. last year where they added less than 18K.  This represented a 37% improvement y/y and 83% improvement sequentially!

On the telco side wireline phone subscriber losses slowed mainly due to TELUS, which had NAL losses of 48k vs. last year’s 55K and Q3 2011 which was 43K losses.  Bell’s NAS losses were worse by 62% y/y whereas TELUS had a good quarter with NAL losses slowing by 12% y/y for the same quarter.  Bell Aliant almost lost as many customers has TELUS, despite having a base that is nearly a million subscribers smaller.  MTS

Q4 2011 wireline phone net adds in Canada

Q4 2011 wireline phone net adds in Canada

In this chart it is clear that the total line lost by the telcos was significantly larger than the total new nets won by the cablecos.  In 2011 this was increasingly the case every quarter and in Q4 more than half the teclo losses did NOT go to cable.  We assume that these customers are going to a combination of wireless (fixed-wireless substitution) which we believe was exaggerated in the GTA and Montreal due to Public mobile.  Some of these customer also went to resellers and VoIP providers.

Telco losses have outstripped cableco gains every quarter since Q4 2008 and this trend is accelerating, as you can see in the next chart.  Bell alone lost more customers than Rogers, Shaw, Videotron and Cogeco together gained.

Substitution = cable gains less teclo losses

Substitution = cable gains less teclo losses

The chart below still shows how Bell and TELUS dominate the landline space, mainly due to their enterprise dominance.  We expect that this will accelerate down with cheaper and better VoIP solutions.  Bell is already on a downward trend, although much of this is in consumer.

Q4 2011 landline subscribers

EOP landline

In wireline phone, the telcos are losing customers at a faster rate than cablecos are gaining them.  While the overall residential market is in decline, cablecos are still gaining share.   Shaw has passed the 40% of share between them and TELUS, Rogers+Videotron+Cogeco now have 45% of the markets that Bell operates in and we believe that Eastlink (which is a private company, so does not report these numbers) is over 50% in urban areas.

Q4 2011 Residential wireline phone subscribers

Q4 2011 Residential wireline phone subscribers

In summary, after many quarters of cable companies easily picking up phone subscribers, their growth has slowed but in aggregate, the telcos continue to lose customers.  Since 2007 cable companies have gained 144,000 customer per quarter and on average the telcos have lost 164,000 customers per quarter.  (c) Alphasynb


No comments