While we take some time to digest the firehose of announcements around Mobile World Congress, I thought it would be useful to review some key statistics about the current state, shape and size of the global mobile network industry.
The Big Picture
GSMA reports 7.7 Billion connections and 3.8 Billion unique subscribers.
Operator revenues exceed $1 Trillion, giving an average monthly rate of $10.
Cisco VNI reports a much higher 4.8 billion users, but almost aligned with the 7.9 total connections.
Overall global subscriber growth isn’t that high (around 5% per annum), so we’re unlikely to suddenly see another billion users join over the coming year.
Traditional fixed voice lines disappearing, fixed broadband growing
Total traditional fixed voice lines have been in decline since it peaked at 1.3 billion in 2008, with significant reductions in developed countries. European rates (as percentage of population, not houses or offices) dropped from 45% to 26% in the last 10 years. US penetration (as a percentage of households) dropped from 95% to just over 50% in a similar period – it won’t be long before cell phone only households predominate. However, this masks a strong takeup in VoIP from cable TV broadband providers and others which now comprise over 20% of the total market. See graph below of Global market split from Telegeography.
Fixed broadband worldwide grew from 220 million to 794 million in that same 10 year timeframe.
Analysis Mason observe that year-on-year revenue for both fixed and mobile has converged and returned to positive levels, albeit just 0.5% across Europe overall. (Mobile revenues had been declining at a rate of 5% just two years ago).
Mobile Ad spending becoming significant, growing very rapidly
Compared to the $1 trillion mobile operator revenues, mobile Ad spending of $100 billion may seem comparatively small. US and China dominate with 62% of that total. Analysts differ on whether 2016 or 2018 will be the year when mobile Ad spend overtakes desktop, but it’s likely to be not far off. US Mobile Ad spending alone is $30 billion, growing by 38% to $42 billion in 2016 according to eMarketer.
CAGR (Compound Annual Growth Rate) is forecast to be 35.6% for Mobile advertising, higher than M2M (16%, Smart Building Services (e.g. NEST) 22% or Cloud Infrastructure Services (10.9%).
Mobile payments is also growing rapidly, but the transaction revenues are small (Apple gets 0.15% when using ApplePay). Smartphone contactless payments amounted to $8.71 billion in 2015, are forecast to increase to $27 billion in 2016 and $210 billion by 2019 [eMarketer]. That matches my observation that I already spend far more using ApplePay than I spend on my mobile phone subscription.
The mix of 2G, 3G and 4G
Those of us fortunate to live in developed countries might think that most of the world has adopted LTE already. That’s far from the case. 2G GSM remains the dominant technology worldwide.
Over the past year, 2G is down 10% (3.7 billion), 3G is up 19% (2.1 Billion) and 4G is up 130% (0.9 Billion)
These declining 2G figures may suggest to you that it might be the first of the generations to be killed off. Globally, peak 3G is forecast to be around 2020, when LTE will just have caught up with it, sharing 40% each and relegating 2G to around 14%. There are still many devices embedded with 2G and it’s important for the M2M market.
Cisco VNI presents the information in an even more stark contrast. LPWA is Low Power Wireless Access, such as Sigfox, LORA and many other competing standards, used by the Internet of Things.
GSM starts disappearing within a year
Nonetheless, AT&T plan to switch their 2G service off in less than a year, on 1st Jan 2017. Many of their 17 million M2M customers still use it, although there’s been plenty of warning and no new provisioned devices for many years. Telstra Australia will also switch off on the same day, claiming GSM currently only handles 1% of traffic.
For those of us who were around at the launch of GSM, it’s a major milestone as we say goodbye to an old friend.
Whereas other countries plan to kill 3G first
Elsewhere, Telenor (Norway) announced last year they would close down 3G first (in 2020), with GSM continuing on to 2025. I think that’s more likely to be the common approach in Europe and elsewhere. Modern macrocell basestations can be remotely software configured to use different radio technologies, avoiding waste in expensive capital assets.
Ovum think that 2G and 3G will start to disappear from 2020 in the UK. I have my doubts about turning off 2G completely. Several UK operators have told me they plan to keep a sliver of 2G alive for quite some time, so I suspect 3G will slowly recede in Europe over the next 10 years.
LTE take-up is happening quickest where you least expected
North America boasts the highest percentage take-up of LTE, with 51% of subscriptions (as of Sep 15) compared to Europe’s paltry 21%. China Mobile wins on sheer numbers though, with 312 Million compared to North America total of 250 Million (as of end Dec 15).
What can be easily overlooked is the recent rapid rise of TD-LTE which comprises around 40% of all LTE devices. It’s predominant in China but also present in Japan, and soon to be launched by Reliance Jio in India. Over 70 networks are now commercially live with TD-LTE, and 21 have both FDD and TDD modes running in parallel (source: GSMA).
Ericsson’s regional bar chart shows likely changes in technology mix over the next five years. Investments across Europe should bring LTE coverage up to match North America by 2020.
The transition to smartphones
As part of the transition to 3G and LTE, there’s been a rapid adoption of smartphones which now comprise 50% of all cellular devices. Today’s 3.3 Billion smartphones will grow to 6 Billion by end 2020, raising the proportion somewhere between 67% (Ovum) and 79% (Cisco).
In the US market, 65% of all mobile phones are smartphones and comprise 84% of new sales. Apple sales are 39% market share (down from 47% last year), and we'll leave it to others to assess the longer term direction of Apple iOS and Android in the smartphone market. They do seem to be increasing their dominance, now over 95%.
VoLTE needed to refarm old spectrum
A key driver for running voice service over LTE is to enable it to replace 2G and 3G service. That needs both newer smartphones (above) and the VoLTE voice service to work. The latter has taken quite some time to roll out – it’s complex, has lots of interfaces and little extra subscriber revenue has materialised.
Only 40 mobile networks had deployed VoLTE by the end of 2015. Some 224 smartphone devices support the service. Some operators are further ahead – T-Mobile USA reports 40% of voice calls are VoLTE - but I still see relatively few VoLTE interconnections between networks.
Amdocs report that VoLTE remains unreliable, with 4x or 5x call drops of the more mature 2G and 3G systems. [I clarified from them that this is not the case for mature, well engineered deployments, but can occur in less well prepared networks during the earlier stages of rollout].
VoLTE deployment is also important to open up use of LTE only small cells. Where voice is predominantly VoLTE, then it makes sense. Meanwhile, I'd see continuing demand for 3G and 3G/LTE multimode product.
You might think that voice minutes are all migrating to OTT providers (Skype, Facebook, Facetime etc). Ovum estimate that to be only 8% at the moment, warning that competition is becoming more intense. US mobile voice calls halved in duration between 2005 and 2012, as we switch to email and messaging for much of our communication.
We’re all watching an awful lot more video than before, not broadcast TV but video clips on YouTube and Facebook. This remains by far the highest proportion of traffic and is only set to increase.
In 2015, video accounted for 50% of mobile data traffic and is set to grow to 70% by 2021. The compound growth rate (CAGR) is 55% during this period.
Average smartphone traffic is 929Mbytes/month now, growing to 4.4GBytes by 2020.
Small Cell Growth
It would seem churlish to exclude a forecast of Small Cell growth during this period. I believe much video is viewed when users are stationary and indoors, and that a greater proportion will be transferred to that method over the coming five years. The Enterprise sector is definitely the one to watch.
Here are two independent Small Cell Analyst forecasts, as presented in our recent Analyst Spotlight webinar, from Caroline Gabriel and Joe Madden respectively.