www.analysysmason.com

ANALYSYS MASON QUARTERLY

Welcome p3

5G will require new as well as established spectrum bands, but the availability of new bands is not confirmed p4

Active RAN sharing business models can bring benefits to towercos as well as operators p5

The emergence of the next TV distribution platform in Europe p6

Pricing for data profitability: usage-based segmentation can identify profit drains p7

Digital economy opportunities in AsiaPacific and the Middle East and Africa: key verticals p8

Bitstream and VULA tariffs in several European countries will need rebalancing to enable IPTV replicability p9

The digital switchover in the next 5 years is an opportunity to rethink the TV market in many countries p10

Content and application providers contribute over USD30 billion per year to internet-related facilities and networks worldwide p11

Inside

OCT-DEC 2014

3Welcome to the Analysys Mason Quarterly.

WELCOME

The fourth edition of Analysys Mason Quarterly for 2014 covers several topics that are trending in the telecoms industry, including 5G, data monetisation and digital economy opportunities in emerging markets.

The process of defining the next generation of wireless communications 5G is now underway. In our first article, Janette Stewart looks at the progress towards a global vision for 5G, and how a key aspect of 5G will be to integrate the various approaches and bands within a harmonised global framework. Tackling a more immediate issue, Amrish Kacker demonstrates how usage-based segmentation can help mobile operators to identify profit drains and improve data monetisation.

We have three articles focusing on developments in the TV market. Matt Yardley considers the potential for broadband networks to become the main TV distribution platform, while Omar Bouhali looks at how the emergence of IPTV may require incumbents to revisit bitstream and VULA pricing. Finally, Llus Borrell provides his view on the future evolution of the digital switchover based on his analysis of developments across different countries and regions worldwide.

On emerging markets, we have an article that examines the digital economy opportunities in AsiaPacific and the Middle East and Africa, based on data from Analysys Masons Digital Economy Readiness Index (DERI).

We welcome the opportunity to discuss your views on these and any other key industry topic. I look forward to hearing from you.

Bram Moerman CEO

The process of defining the next generation of wireless communications (5G) is now underway, and the economic impact of decisions that national governments will make about the assignment and award of spectrum for 5G is potentially significant. There is already immense interest and lobbying from a wide range of stakeholders in relation to the development of 5G, and the associated market, technology and spectrum issues, even though 5G networks are unlikely to be needed until after 2020 in many markets. A key trend is the potential use of technologies deployed in the millimetre wave portion of radio spectrum. This article considers work in progress towards achievement of a global vision for 5G, and what the spectrum implications might be.

Mobile operators are currently investing heavily in 4G networks, and this investment is likely to continue well beyond 2020. New 4G features such as carrier aggregation make it possible for mobile operators to aggregate two or more radio frequency carriers to boost user-data throughput. Better ways of using spectrum such as adding supplementary downlink bands to cater for significant demand for data download will also improve utilisation of spectrum in many markets. The need to monetise LTE investments is already creating shifts in mobile markets and business models, with a greater focus on sharing between different network elements, network consolidation and new pricing models being launched for mobile data services.

However, despite these advances, 4G networks will not offer a serious challenge to fibre broadband during the next decade and are unlikely to achieve the speeds, coverage, reliability and performance required in future wireless networks. This is partly due to the economics of 4G network build as well as for technological reasons. One of the key motivators for 5G is thus to provide ubiquitous, high-speed, high-quality wireless broadband coverage to meet societal and industrial needs beyond 2020. This positions 5G at the core of many national and regional government targets on next-generation broadband availability and use, such as the Digital Agenda for Europe.

Research programmes, industry collaborations and standardisation debates now underway to define a 5G global vision will ultimately determine the technological, network and spectrum requirements of 5G. Consensus-building and harmonisation are anticipated between 2015 and 2020 ahead of full 5G commercial launches (see Figure 1).

At this early stage, a diverging range of business models, applications, market sectors and devices are emerging. Possible technological routes to 5G include entirely new radio technologies and techniques, as well as further evolution and deployment of existing technologies (LTE-A and Wi-Fi, for example).

A key difference between 5G and earlier generations of mobile technology is that the focus of research is on finding the best techniques to improve spectrum utilisation (i.e. bits per Hertz per unit area), rather than on improving spectrum efficiency (i.e. bits per Hertz). This is because improvements in spectral efficiency are constrained by background noise, meaning that improvements through coding and modulation design become more difficult and less effective (the Shannon Limit). However, new technological approaches can substantially improve spectrum utilisation, defined in terms of bits/Hertz/cell (or area). They can also enable networks to become more flexible and suitable to carry some of the new applications and use cases being foreseen within 5G.

Proposed technologies such as massive MIMO, super-dense meshed cells and macro-assisted small cells (phantom cells) are all possible 5G radio access techniques targeting better spectrum utilisation, higher speeds and lower latency. From the users perspective, the aim is to provide a better and more-consistent service regardless of location.

Many of the technologies being researched are inherently better suited to being deployed in very high frequency bands in the millimetre range of radio spectrum (current 5G research includes trials conducted in bands such as 15GHz, 28GHz, 60GHz and

70GHz, for example substantially higher bands than mobile communications has traditionally used). This spectrum can better support the use of multiple, miniaturised antennas (since the wavelength of higher frequency bands is shorter and antenna spacing is based on wavelength, so more antenna elements can be accommodated). Furthermore, substantially more bandwidth is available in these bands than in the bands below 1GHz, which is beneficial for providing much wider channels and higher speeds as envisaged by 5G, without the need for multiple antennas.

However, millimetre-wave bands do not lend themselves to providing wide area coverage for mobile devices (and coverage will be essential for some envisaged 5G services, such as IoT applications such as for the automotive industry). Therefore, further spectrum below 1GHz is expected to be needed in many countries to improve mobile broadband coverage.

The spectrum needs for 5G might therefore encompass a range of existing and new bands, which potentially span a wide section of radio spectrum. Different bands will serve different purposes and a key aspect of 5G will be to integrate the various approaches and bands within a harmonised global framework. Early indications suggest that spectrum sharing is likely to be used in a far greater way, which may signal an end to further spectrum being reserved for exclusive mobile broadband use as 5G is introduced.

Analysys Mason has conducted world-leading research into 2G, 3G and 4G networks and continues to be at the forefront of technology and spectrum consulting with 5G.

For more information, please contact Janette Stewart, Principal, at janette.stewart@analysysmason.com

4

ANALYSYS MASON QUARTERLY OCT-DEC 2014

JANETTE STEWART Principal, Consulting Division

5G WILL REQUIRE NEW AS WELL AS ESTABLISHED SPECTRUM BANDS, BUT THE AVAILABILITY OF NEW BANDS IS NOT CONFIRMED

Figure 1: Timeline towards 5G [Source: Analysys Mason, 2014]

2008

LTE introduction(3GPP Release 8)

Global projectsunderway for

SG vision

2010

First LTE-A standard(3GPP Release 10)

2014

3GPPRelease 12

ITU WRC-15

20152020

ITU reporton IMT vision

Spectrumlicensing

Technologyconsensus

3GPP study for 5G(Release 14 onwards)

Beyond2020

Networkimplementations

ITU reporton future

technologytrends

Spectrumharmonisations

WRC-18

Active RAN sharing is likely to be the next significant evolutionary step in infrastructure sharing, unlocking even greater capex and opex efficiencies than passive RAN sharing. Each stakeholder can quickly benefit from those efficiencies through the rapid adoption of optimised business models.

Independent tower companies (towercos) have paid billions of dollars to acquire passive infrastructure from mobile network operators (MNOs). Currently, the success of the towerco business model is predicated on tenancy ratio growth that is, growth in hosting antennas. If MNOs choose to share antennas in an active RAN sharing solution, where does that leave the towercos? This article examines the various kinds of RAN sharing architecture, explores different RAN sharing business models and seeks a win-win market scenario that will ensure benefits for all stakeholders.

Different kinds of active infrastructure sharing vary in terms of the degree of sharing

3GPP has defined and ratified different kinds of architecture with varying degrees of sharing.

Multi-Operator RAN (MORAN) is the simplest scenario, in which only equipment is shared

Multi-Operator Core Network (MOCN), in which both spectrum and equipment are shared

Gateway Core Network (GWCN) is where both the RAN and some elements of the core network are shared.

Sharing can ensure more-efficient usage of spectrum

Antenna and radio technology has progressed rapidly in recent years, such that a single antenna can serve multiple bands between 700MHz and 2.6GHz, although multiband antennas can compromise network performance. However, software-defined radio (SDR) can provide significant benefits by dynamically allocating and aggregating spectrum across multiple bands to multiple service providers, therefore cost-effectively maximising the usage of what can be scarce and costly spectrum.

Appetite and opportunity for RAN sharing differs between European and African operators

An almost 30-year-old legacy of European mobile market fragmentation can make some stakeholders reluctant to explore RAN sharing. However, European operators, faced with flat revenue and forced investment to support burgeoning data traffic demand, are motivated to explore strategies that can unlock significant cost savings.

In contrast, African operators may be more willing to collaborate and engage in RAN sharing. Cost savings are also important in Africa, but there is an additional benefit of rapid coverage gains, motivated by a need to make mobile communications more affordable and drive penetration rates by accessing more low-ARPU subscribers.

Legacy antenna-driven business models are sub-optimal for active sharing

In a RAN-sharing scenario antennas can be shared by multiple MNOs. In order to future-proof themselves against RAN sharing, towercos should adopt a business model driven by usage, for example, by the number of service providers using a tower, not only by the physical space required. Other physical design factors should also be considered, for example, the increasing use of tower-mounted radio units providing MNOs with the benefits of reduced power consumption and/or increased coverage.

TowerXchange spoke to one African towerco that defines a tenant as a standard amount of space and a standard amount of power. This seems to be the prevalent approach; rental rates are dependent on standard equipment configurations. At the recent TowerXchange Meetup Americas, towercos spoke of adding frequency-specific language to their lease contracts, ensuring that each additional tenant pays a monthly lease whether hanging new equipment or sharing existing antennas.

If towercos can agree mutually beneficial contractual terms with MNOs, they could become even more investible as a result of RAN sharing, because the load and space required on a tower can be less for one shared antenna than for three or four

antennas doing the same job, freeing up finite tower capacity to sell to more tenants.

ROI expectations are satisfied through passive asset transfer

The motivation to transfer infrastructure assets from MNOs to towercos is clear when one considers both the immediate cash release for the MNO and the differing return on investment (ROI) horizon expectations for each party. Specialist infrastructure investors readily incorporate towers into an investment mindset that typically includes transport and mining infrastructure, and might calculate ROI over anything from 10 to 25 years, while MNOs prefer 35-year ROI horizons.

Analysys Mason recently conducted a project to develop the business case and operational model to establish a greenfield towerco. The project considered the potential synergies of augmenting the business of an established fibre-based service provider by offering a portfolio of managed active network services as well as traditional towerco services. However, the challenge was to identify a complementary investment profile to add shareholder value and define the operational demarcation between the two entities considering the different market and operational requirements.

Different investment models are attractive to different stakeholders and skill sets are not easily transferable between active and passive operations. In recent years, MNOs have tended to move away from long-term investments by either setting up separate infrastructure entities, as in the Indian mobile telecoms market, or by divesting infrastructure altogether through a sell-and-leaseback arrangement to independent towercos that have a longer-term investment view.

Analysys Mason is a pioneer of network-sharing solutions, having supported very early passive sharing ventures as well as recent leading active sharing solutions. We have extensive experience of providing valued advice to both towercos and MNOs to optimise their businesses through informed infrastructure and technology investment.

For more information, please contact Mike Pearson, Partner, at mike.pearson@analysysmason.com or Kieron Osmotherly, CEO, TowerXchange, at kosmotherly@towerxchange.com

ACTIVE RAN SHARING BUSINESS MODELS CAN BRING BENEFITS TO TOWERCOS AS WELL AS OPERATORS

ANALYSYS MASON QUARTERLY OCT-DEC 2014

MIKE PEARSON Partner, Consulting Division

5

An interesting and potentially disruptive theme that has emerged from our recent policy work is the potential for broadband (IP) networks to substitute for free-to-air (FTA) digital terrestrial TV (DTT) and satellite networks as the main TV distribution platform. Such a transformation seems likely, but it is uncertain when it will happen. This article examines the factors that will influence when platform substitution could take place.

Increasing high-speed broadband coverage means that the supply-side conditions for platform substitution could be met in 510 years

The European Commission (EC) has launched a major push to stimulate broadband deployment through its Digital Agenda initiative and elements of its recent Connected Continent legislative package. The two initiatives plan to deliver ubiquitous, high-speed networks with assured quality of service. Should these policy goals be met, then some of the necessary conditions for TV distribution platform substitution could be in place within the next 510 years. In some European countries, this could happen even sooner.

In the UK, the government has moved quickly to stimulate broadband deployment beyond the commercially viable footprint. The government aims for 95% of households to have access to broadband at speeds of at least 24Mbit/s by the end of 2017, and for at least 99% of households to be covered by the end of 2018. The speed ambition for this last 45% is not yet defined, but it seems likely that it will be sufficient to at least deliver broadcast and on-demand TV streams of standard definition content. For comparison, these expected coverage levels exceed that for commercial DTT multiplexers in the UK at 90%, and are close to that of public service DTT multiplexers at 98.5%. It is also highly significant that multicast one of the key technologies to reduce the costs of delivering linear

broadcast content over broadband will be available as a wholesale product to communications providers serving 95% of households (and it could be extended beyond that depending on how the final 45% are served).

The UK is currently one of the leading European countries in this respect, but it seems inevitable that other countries will also reach very high broadband coverage levels. It remains to be seen whether the ECs ambition of high-speed networks being available to all EU households by 2020 will be reached.

Demand-side factors may ultimately dictate when IP platforms could substitute for DTT and satellite

The supply of the right infrastructure would not be sufficient to enable such a transformation in itself. The most-popular FTA content would also be needed, and public service broadcasters and other major content owners will need to consider the implications of making this linear content available on an IP platform. However, growing consumer demand for watching both linear and non-linear content on multiple devices, and the growing trend for the main household TV being connected to the Internet, suggest that content will ultimately need to be made available on all platforms (including IP) in equivalent forms. It is also evident that many leading content providers are already heading in this direction, for example, by experimenting with new approaches to delivering high-profile events over IP. Furthermore, hybrid DTT-IP and satellite-IP models are appearing that demonstrate that the DTT and satellite communities recognise the importance to users of the broadband-enabled experience.

It may be that demand-side factors ultimately dictate the timing of when IP platforms could fully substitute for DTT and satellite, or indeed if hybrid models exist in the long term. The propensity to connect the

main household TV to the Internet will vary significantly by user segment, and will be affected by the installed base of TVs, set-top-boxes and in-home networks. Inertia will also undoubtedly play a part for some people. For the elderly and disadvantaged groups, it seems likely that FTA DTT in particular will remain important for a long time, because broadband take-up among these groups is currently low in most countries. However, many governments have already identified this issue as a priority getting those who do not currently use the Internet online is critically important for social engagement and reducing the cost of delivering public services in the future. Hence, most of these people will eventually become Internet users.

In summary, we see strong alignment in EU and national policies that, by design or otherwise, will promote the deployment of a pan-European IP-based TV distribution platform as an alternative to FTA DTT and satellite. Aside from the numerous regulatory and licensing and demand questions this raises, the commercial and business model impacts could be transformational, particularly if IP-distribution platforms were to carry all FTA channels and be open to all ISPs at little or no incremental cost. For the traditional TV platforms the issue is not so much when IP platforms would serve most users; rather it is the impact on their business models of even a modest proportion of their users moving over to IP. Hybrid models may help, but their longevity is unknown at this point. The timing of when IP will become mainstream therefore depends on many factors.

Analysys Mason has been closely examining these factors and we welcome the opportunity to discuss with stakeholders what this means for them.

For more information, please contact Matt Yardley, Partner, at matt.yardley@analysysmason.com

6

THE EMERGENCE OF THE NEXT TV DISTRIBUTION PLATFORM IN EUROPE

MATT YARDLEY Partner, Consulting Division

ANALYSYS MASON QUARTERLY OCT-DEC 2014

A problem facing mobile operators is that a small proportion of subscribers consume a disproportionately large proportion of data. Analysys Mason has undertaken a detailed assessment of data profitability based on detailed CDRs for an operator in South-East Asia. Our assessment found that the 5% of subscribers with the heaviest data usage use 30% of total network data consumed, but contribute only 6.5% of revenue and a mere 1% of overall margins (defined as revenue less network costs and subsidies), as illustrated in Figure 2. The cumulative margin line has points of decrease indicating unprofitable users in all usage segments. Notably, these are not only in high data usage segments, as is generally believed.

Usage-based segment heat maps can identify profit drains

Usage-based segment heat maps provide an overview of the level of profitability for different customer segments, as illustrated in Figure 3. Typically, higher-ARPU customers provide higher margins, although some lower-ARPU segments can also be highly profitable (these tend to be voice-centric users). However, in order to manage data profitability, operators need to assess how best to manage the profit drains that is, segments that have negative margins and how to improve margins for the low-margin users. In the example shown in Figure 3, just eliminating negative margin users will improve total margins by 2.2% even though revenue will decline by 3%.

Managing segment profitability can ensure profitable data growth in three ways

Usage segmentation will provide operators with the basis for ensuring three key areas of profitable data growth.

Eliminating subscribers who appear to be profit drains, based on heat maps. A time-of-day usage assessment for profit drains can provide significant additional analysis.

Reduction of subsidies or promotions for unprofitable customers. A customer lifetime value assessment needs to be carried out after usage segmentation.

Improving margins for existing customers. This could be achieved by setting tiers for data bundles based on growth of data usage by segment.

Analysys Mason supports operators in developing robust strategies to improving data profitability, including using a structured approach to data costing (see our article Mobile data services: a structured approach to data costing is the first step to profitability).

For more information please contact Amrish Kacker, Partner, at amrish.kacker@analysysmason.com or Andy Leonardi, Consultant, at andy.leonardi@analysysmason.com

PRICING FOR DATA PROFITABILITY: USAGE-BASED SEGMENTATION CAN IDENTIFY PROFIT DRAINS

AMRISH KACKER Partner, Consulting Division

7

ANALYSYS MASON QUARTERLY OCT-DEC 2014

Figure 2: Distribution of data usage, revenue and margin, South-East Asian operator [Source: Analysys Mason, 2014]

Figure 3: Customer segment margin heat map (after network costs/subsidies), South-East Asian operator [Source: Analysys Mason, 2014]

20%

0%0%

Cumulative data used

20% 40% 60% 80% 100%

40%

60%

80%

100%

Percentage of subscribers

Per

cent

age

cum

ulat

ive

subs

crib

ers

Cumulative revenue Cumulative margin

70%

Increasing ARPU

Incr

easi

ng d

ata

usag

e

Margin

AsiaPacific (APAC) and the Middle East and Africa (MEA) are two of the most fertile regions for digital economy deployments. Data from Analysys Masons Digital Economy Readiness Index (DERI) a compilation of over 340 digital economy initiatives by 32 of the largest operators worldwide in terms of mobile revenue reveals a divide between these two regions, as well as potential opportunities for operators and vendors.

Each initiative in the DERI receives a readiness score assessed across two distinct axes: the scale or type of an initiative which indicates its size and reach and its maturity, which indicates its position within the lifecycle of a digital economy initiative

APAC and MEA have distinct digital economy profiles

Data from the DERI enable a top-level view of each regions distinct digital economy profile, and the key verticals in each of them:

The MEA region has a high number of mobile health deployments. A third of the initiatives that we tracked in MEA were related to mobile health, making this the vertical with the largest number of deployments in the region. In contrast, mobile health represented 18% of initiatives in APAC. The relatively high level of activity in mobile financial service initiatives in APAC indicates that innovation may also be happening with smaller, local operators and not only the largest players.

Mobile education is important in both regions. 46% of all mobile education initiatives tracked in DERI were observed in the APAC region. However, MEA leads in terms of average readiness score. Education is perceived as a key social mobility enabler in these markets and can be effectively monetised directly (subscription services), or indirectly (churn prevention).

APAC dominates MEA in terms of the number of cloud-based services initiatives.This is stimulated by smartphone penetration Analysys Mason expects smartphone penetration as a proportion of total active handset connections in APAC to reach 34% by the end of 2014, while it will be 16% in MEA.

The DERI assesses digital economy initiatives on more than just numbers of deployments

Operators in MEA, such as Ooredoo, MTN, Orange and Saudi Telecom, have deployed fewer mobile health and cloud-based service initiatives than operators, such as NTT Docomo and SoftBank, in APAC. However, the initiatives in MEA receive higher readiness scores on average.

In MEA, mobile education and mobile financial services are the two most successful verticals in terms of average readiness score (see Figure 4). MEA commands the highest scores for mobile financial services worldwide, followed by APAC. In the former, Vodafones M-Pesa, MTN Mobile Money and Orange Money are best-practice examples of productmarket fit. APACs deployments have been driven by the success of contactless and mobile-only payments, and money solutions. Nevertheless, average readiness scores for mobile financial services are slightly lower in APAC than in MEA, because the most successful initiatives are often limited to Japan or South Korea, which are not representative of the whole region.

Opportunities exist in verticals that have a high average readiness score and a small number of deployed initiatives. In APAC, smart homes, mobile education and mobile agriculture are potential growth verticals for

mobile operators. In MEA, opportunities exist in mobile education, mobile financial services, and mobile health.

Analysys Masons Digital Economy Readiness Index (DERI) presents a view of operators ability to capitalise on emerging digital economy verticals, quantifying the readiness of their initiatives in:

mobile financial services

mobile health

mobile commerce and advertising (including location-based services)

mobile education

cloud-based services (including OTT, mobile content, big data, and software)

mobile agriculture

smart homes

venture capital or start-up incubator/accelerator programmes.

The DERI integrates data from each deployment across a number of categories, including vertical, scale, maturity, target customer segment, countries and regions. Each initiative in the DERI receives a readiness score using Analysys Masons proprietary methodology.

For more information, please contact Enrique Velasco-Castillo, Analyst, at enrique.velasco-castillo@analysysmason.com

DIGITAL ECONOMY OPPORTUNITIES IN APAC AND MEA: KEY VERTICALS

ENRIQUE VELASCO-CASTILLO Analyst, Research Division

8

ANALYSYS MASON QUARTERLY OCT-DEC 2014

Figure 4: Number of initiatives and average readiness score by vertical, AsiaPacific and Middle East and Africa [Source: Analysys Mason, 2014]

0

0 5 10 15 20 25 30

Aisia-Pacific

50

45

40

35

30

25

20

15

10

5

Number of initiatives

Ave

rage

rea

dine

ss s

core

Middle East and Africa

Key: APAC = Asia Pacific; MEA= Middle East and Africa.

Cloud-basedservices

Mobile agriculture

Mobile commerceand advertising

Mobile education Mobile financial

services

Mobile health

Smart homes

Cloud-basedservices

Mobile agriculture

Mobile education

Mobile financial services

Mobile health

Smart homes Venture capital accelerator

Venture capital accelerator

Bitstream and VULA are wholesale offers that many alternative operators around the world use to provide broadband offers in regions where they do not identify a business case for deploying their own local loop or for using physical access. The development of IPTV, in particular for non-linear content, will require several incumbent operators to revisit the pricing structures of these offers, because traffic-driven price components may make it impossible for alternative operators using bitstream or VULA to replicate some of the incumbent operators offers.

IPTV can have a significant impact on broadband average traffic profile

Bitstream and VULA pricing structures generally include a fixed component per connection (that typically depends on connection characteristics, such as peak speed) and a variable component based on traffic (typically measured as peak traffic, 95 percentile traffic or average throughput).

When regulators test whether bitstream or VULA offers enable alternative operators to replicate the incumbent operators offers, they have to use traffic assumptions these can be based on actual market data or on a forecast evolution.

IPTV can have a dramatic impact on such traffic assumptions, because IPTV requires significantly more bandwidth than applications such as web browsing or VoIP. Content delivery networks (CDNs) can limit the traffic increase in some cases (by multicasting linear channels or caching popular content, for example), but they are less efficient when demand becomes a wide diversity of non-linear programmes, and they may be too expensive to address some low-density areas. When IPTV leads to a strong increase in the average traffic profile, bitstream and VULA offers that had been designed to enable replicability of the incumbent operators offers may no longer enable this replicability because of the

increase in traffic-driven costs. This can be the case even when additional revenue that the incumbent operator can get from these IPTV services is taken into account (see Figure 5).

Ensuring replicability with IPTV is likely to result in a decline in the variable component of bitstream and VULA prices

A recent Analysys Mason report showed that high usage had made the bitstream and VULA services with bandwidth-based backhaul charges very uncompetitive in several European countries.

In its latest Market 5 analysis (decision no. 2014-0734, 26 June 2014), French regulator ARCEP imposed an obligation on Orange to adapt its bitstream offer in zones where it is the only wholesale broadband service provider, by taking into account the development of non-linear IPTV services during the next 3 years in the cost accounting allocations used to set these bitstream tariffs. This review should lead to a strong decline in the traffic-driven component, which is necessary to enable alternative operators to replicate Oranges IPTV services.

In most countries where the development of non-linear IPTV services leads to a significant change in average traffic profile, regulators and operators will need to revisit the pricing structure of bitstream and VULA offers in order to ensure replicability. This is likely to result in a reduced variable component of bitstream and VULA prices and, potentially, in an increased fixed component.

Analysys Mason regularly supports regulators and operators in markets analyses and in adapting regulated offers to changes in the competitive environment.

For more information, please contact Omar Bouhali, Principal, at omar.bouhali@analysysmason.com

BITSTREAM AND VULA TARIFFS IN SEVERAL EUROPEAN COUNTRIES WILL NEED REBALANCING TO ENABLE IPTV REPLICABILITY

OMAR BOUHALI Principal, Consulting Division

9

ANALYSYS MASON QUARTERLY OCT-DEC 2014

Figure 5: Impact of a strong increase in the average traffic profile on replicability tests [Source: Analysys Mason, 2014]

Incumbentoperator retail

offer

Replicability testfor alternative

operator

Margin

Marginsqueeze

Limited IPTV usage

Incumbentoperator retail

offer

Replicability testfor alternative

operator

Significant non-linear IPTV usage

Mon

thly

rev

enue

cos

ts p

er s

ubsc

ribe

r

Retail ARPU Fixed component Traffic-driven component Retail costs

The deadline for the digital switchover (DSO) from terrestrial television has been set for between 2015 and 2020 in many countries worldwide by ITU and national authorities. The DSO will be used as an opportunity to rethink the TV market in many countries during the next 5 years. The approach governments and regulators will take will depend on their aims and objectives, which the private sector will also need to consider.

These objectives are typically among the following:

supporting digital TV transformation and the launch of new services (HDTV and HbbTV, for example)

ensuring that comprehensive and universal TV services are affordable to all

opening up the TV market to new entrants, therefore diversifying the TV offering to consumers

freeing up spectrum for mobile broadband and to obtain significant revenue

supporting national original content and independent production

supporting national industrial policy.

Many articles on DSO have been published that focus on the associated narrower operational and technical decisions but fail to address these key objectives. In our experience, while these operational decisions are crucial, the success of DSO also relies on three main issues associated

with these objectives: the attractiveness of digital terrestrial TV (DTT) content offers, the appropriateness of the regulatory regime and the ability to enforce the regulatory regime.

Analysys Masons worldwide experience in DSO projects illustrates the diversity of situations and policy decisions

We have worked on DSO projects across all regions worldwide. The fundamental issues and analyses are the same, but diverse situations, policy decisions and private investments lead to different outcomes for the TV market. Below we outline some high-level differences to illustrate this diversity and examine the key issues in each case.

Europe launching new services with spectrum issues at the centre of the debate. Much of the focus has been on the associated UHF broadcasting spectrum digital dividend bands that could be migrated to mobile use (initially 800MHz and now potentially 700MHz as well). In markets where DTT content offerings have not improved significantly, the platform has struggled to remain competitive. More widely, the TV industry focuses on making DTT competitive by moving to HDTV and UHDTV, adopting non-linear TV standards like HbbTV, and ensuring that DTT is available on mobile devices.

Africa rethinking the TV market with a focus on local content and network investments. The situation in Africa highlights different technical challenges and issues. Although the situation varies between countries, analogue terrestrial infrastructure can be less universal than in other regions, meaning that satellite or cable are taking some lead. Therefore, DSO policy makers need to focus on their DTT content strategies as well as on taking action to improve the terrestrial distribution reach.

Latin America diverse industrial policy, rethinking the TV market and affordability. DSO is planned to take place between 2015 and 2020 in most countries in the Latin America region. In terms of industrial policy decisions, we have seen a variety of choices influenced by different models, such as those of the USA, Europe and Japan, both in the selection of technical standards (ATSC, DVB and ISDB) and the structure of the TV market.

AsiaPacific offering attractive and affordable TV to all. In the AsiaPacific region, DSO is planned for most countries between 2015 and 2020. The main challenge is similar to that in other regions to ensure that DTT is accessible and attractive to all.

USA broadcast spectrum auctions are the central focus. The focus in the USA has mainly been on spectrum release for mobile use after the associated auction. Following the 700MHz auction in 2009, plans are underway for an incentive auction of spectrum below 698MHz during the next 12 months.

We expect significant progress towards DSO in many countries worldwide during the next 5 years that will lead to changes in the TV markets.

Through its team and network of offices, Analysys Mason has international and local TV experience in more than 35 countries worldwide, including DSO and all the associated issues.

For more information, please contact Llus Borrell, Partner, at lluis.borrell@analysysmason.com

10

THE DIGITAL SWITCHOVER IN THE NEXT 5 YEARS IS AN OPPORTUNITY TO RETHINK THE TV MARKET IN MANY COUNTRIES

LLUS BORRELL Partner, Consulting Division

ANALYSYS MASON QUARTERLY OCT-DEC 2014

The Internet is often abstracted as a cloud, but in practice it relies heavily on bricks, mortar, cables and computer boxes the intangibility and ubiquity of the cloud is built on real, physical infrastructure.

Many different market participants invest in the networks, facilities and equipment that make up the Internet: Internet backbone providers; Internet access providers; content and application providers; and a range of specialised service providers. This article examines the investments of content and application providers.

Our study found that content and application providers are major investors in the networks that make up the Internet

In a recent study, we focused on the investment that content and application providers make in Internet infrastructure.1 These investors include pure online companies such as Facebook, Google and Spotify, but also the online businesses of multi-platform players such as the BBC or The New York Times. We found that these players are major investors in Internet infrastructure, from data centres to submarine cables and the multitude of servers that store, process and serve content to end users.

This investment relates to three major activities.

Hosting of content and data this includes the storage devices and servers that reside in large data centres located around the world.

Transport includes the high-capacity fibre-optic cables (terrestrial or submarine) that are used to carry content from the hosting location to the edge of the Internet access providers network.

Delivery includes the equipment necessary to hand over the content to the Internet access provider, which then delivers it across the middle and last mile to the end user. This includes in particular equipment used in Internet exchanges, private peering points, and content delivery networks (CDNs).

Based on this categorisation, we found that the investment attributable to content and application providers was approximately USD33 billion per year during 20112013 (this includes direct investment as well as investment from third parties such as backbone providers or data centre operators that support and are sold to content and application providers). Investment is growing rapidly as demand for content and applications delivered over the Internet increases this investment grew by 13% per year between 2011 and 2013. Figure 6 summarises these investments.

This level of investment is significant, particularly because it is additional to the investment that content and application providers make in their core businesses, which may be software, digital content or ecommerce.

We also found that Europe appears to be the largest destination for this investment. This region is a hub for Internet traffic, where many international cables meet, it hosts the worlds largest IXPs, and has a large

population of Internet users. This is attracting investment by US companies, particularly in data centre facilities, as well as investment by local content and application providers such as the BBC and Spotify.

The Internet relies on ever more complex infrastructure, which encompasses data centres, transmission networks, connectivity as well as access. The source of investment in this infrastructure is evolving, and content and applications Providers are responsible directly or indirectly for a material and constantly increasing amount.

Analysys Mason works across the telecoms, media and Internet value chains to deliver research, analysis and strategic recommendations to some of the largest companies in the world, policy makers, governments and investors.1For more information, see Analysys Masons Investment in networks, facilities and equipment by content and application providers. Available at www.analysysmason.com/About-Us/News/Press-releases1/Internet-players-invest-USD100-billion-in-the-physical-Internet-over-three-years.

For more information, please contact Andrew Kloeden, Principal, at andrew.kloeden@analysysmason.com or David Abecassis, Principal, at david.abecassis@analysysmason.com

Andrew Kloeden Principal, Consulting Division

CONTENT AND APPLICATION PROVIDERS CONTRIBUTE OVER USD30 BILLION PER YEAR TO INTERNET-RELATED FACILITIES AND NETWORKS WORLDWIDE

11

ANALYSYS MASON QUARTERLY OCT-DEC 2014

Figure 6: Approximate annual investments by content and application providers, 20112013 [Source: Analysys Mason, DatacenterDynamics, TeleGeography, Informa, company data, news reports, PeeringDB, comScore, Sandvine, 2014]

Hosting Submarine

29.60

0.92 1.730.52 0.29

Terrestrial Peeringpoints

Content deliverynetworks

Inve

stm

ent (

USD

bill

ion)

Mid-point Minimum Maximum

40

35

30

25

20

15

10

5

0

www.analysysmason.com