Why the consumer IoT is stuck in the slow lane

A slow start for NB-IoT and LTE-M

For telcos around the world, the Internet of Things (IoT) has long represented one of the most promising growth opportunities. Yet for most telcos, the IoT still only accounts for a low single digit percentage of their overall revenue. One of the stumbling blocks has been relatively low demand for IoT solutions in the consumer market. This report considers why that is and whether low cost connectivity technologies specifically-designed for the IoT (such as NB-IoT and LTE-M) will ultimately change this dynamic.

NB-IoT and LTE-M are often referred to as Massive IoT technologies because they are designed to support large numbers of connections, which periodically transmit small amounts of data. They can be distinguished from broadband IoT connections, which carry more demanding applications, such as video content, and critical IoT connections that need to be always available and ultra-reliable.

The initial standards for both technologies were completed by 3GPP in 2016, but adoption has been relatively modest. This report considers the key B2C and B2B2C use cases for Massive IoT technologies and the prospects for widespread adoption. It also outlines how NB-IoT and LTE-M are evolving and the implications for telcos’ strategies.

This builds on previous STL Partners’ research, including LPWA: Which way to go for IoT? and Can telcos create a compelling smart home?. The LPWA report explained why IoT networks need to be considered across multiple generations, including coverage, reliability, power consumption, range and bandwidth. Cellular technologies tend to be best suited to wide area applications for which very reliable connectivity is required (see Figure below).

IoT networks should be considered across multiple dimensions

IoT-networks-disruptive-analysis-stl-2021
Source: Disruptive Analysis

 

Enter your details below to request an extract of the report

The smart home report outlined how consumers could use both cellular and short-range connectivity to bolster security, improve energy efficiency, charge electric cars and increasingly automate appliances. One of the biggest underlying drivers in the smart home sector is peace of mind – householders want to protect their properties and their assets, as rising population growth and inequality fuels fear of crime.

That report contended that householders might be prepared to pay for a simple and integrated way to monitor and remotely control all their assets, from door locks and televisions to solar panels and vehicles.  Ideally, a dashboard would show the status and location of everything an individual cares about. Such a dashboard could show the energy usage and running cost of each appliance in real-time, giving householders fingertip control over their possessions. They could use the resulting information to help them source appropriate insurance and utility supply.

Indeed, STL Partners believes telcos have a broad opportunity to help coordinate better use of the world’s resources and assets, as outlined in the report: The Coordination Age: A third age of telecoms. Reliable and ubiquitous connectivity is a key enabler of the emerging sharing economy in which people use digital technologies to easily rent the use of assets, such as properties and vehicles, to others. The data collected by connected appliances and sensors could be used to help safeguard a property against misuse and source appropriate insurance covering third party rentals.

Do consumers need Massive IoT?

Whereas some IoT applications, such as connected security cameras and drones, require high-speed and very responsive connectivity, most do not. Connected devices that are designed to collect and relay small amounts of data, such as location, temperature, power consumption or movement, don’t need a high-speed connection.

To support these devices, the cellular industry has developed two key technologies – LTE-M (LTE for Machines, sometimes referred to as Cat M) and NB-IoT (Narrowband IoT). In theory, they can be deployed through a straightforward upgrade to existing LTE base stations. Although these technologies don’t offer the capacity, throughput or responsiveness of conventional LTE, they do support the low power wide area connectivity required for what is known as Massive IoT – the deployment of large numbers of low cost sensors and actuators.

For mobile operators, the deployment of NB-IoT and LTE-M can be quite straightforward. If they have relatively modern LTE base stations, then NB-IoT can be enabled via a software upgrade. If their existing LTE network is reasonably dense, there is no need to deploy additional sites – NB-IoT, and to a lesser extent LTE-M, are designed to penetrate deep inside buildings. Still, individual base stations may need to be optimised on a site-by-site basis to ensure that they get the full benefit of NB-IoT’s low power levels, according to a report by The Mobile Network, which notes that operators also need to invest in systems that can provide third parties with visibility and control of IoT devices, usage and costs.

There are a number of potential use cases for Massive IoT in the consumer market:

  • Asset tracking: pets, bikes, scooters, vehicles, keys, wallets, passport, phones, laptops, tablets etc.
  • Vulnerable persontracking: children and the elderly
  • Health wearables: wristbands, smart watches
  • Metering and monitoring: power, water, garden,
  • Alarms and security: smoke alarms, carbon monoxide, intrusion
  • Digital homes: automation of temperature and lighting in line with occupancy

In the rest of this report we consider the key drivers and barriers to take-up of NB-IoT and LTE-M for these consumer use cases.

Table of Contents

  • Executive Summary
  • Introduction
  • Do consumers need Massive IoT?
    • The role of eSIMs
    • Takeaways
  • Market trends
    • IoT revenues: Small, but growing
  • Consumer use cases for cellular IoT
    • Amazon’s consumer IoT play
    • Asset tracking: Demand is growing
    • Connecting e-bikes and scooters
    • Slow progress in healthcare
    • Smart metering gains momentum
    • Supporting micro-generation and storage
    • Digital buildings: A regulatory play?
    • Managing household appliances
  • Technological advances
    • Network coverage
  • Conclusions: Strategic implications for telcos

 

Enter your details below to request an extract of the report

Why closing Telefonica Digital should make Telefonica more digital (and innovative)

Several different CSP organisation designs for Telco 2.0 Service Innovation

Telefonica is one of the companies that we have analysed in depth in the Telco 2.0 Transformation Index research. In this report, we analyse Telefonica’s recent announcement that it is restructuring its Digital Business unit. We’ll also be exploring strategies for transformation at the OnFuture EMEA 2014 Brainstorm, June 11-12, London.

Telco 2.0 strategy is a key driver of organisation design

We have defined Telco 2.0 and, specifically, Telco 2.0 Happy Piper and Telco 2.0 Service Provider strategies in other reports  so will not focus on the implications of each on service offerings and customer segments here.  It is, however, important to understand the implications each strategy has on the organisation in terms of capability requirements and, by definition, on organisation design – structure, processes, skills and so forth.

As Figure 1 shows, the old Telco 1.0 world required CSPs to focus on infrastructure-oriented capabilities – cost, service assurance, provisioning, network quality of service, and congestion management.

For a Telco 2.0 Happy Piper, these capabilities are even more important:

  • Being low-cost in a growing telecoms market gives a company an advantage; being low-cost in a shrinking telecoms market, such as Europe, can mean the difference between surviving and going under.
  • Congestion management was important in the voice-oriented telecoms market of yesteryear but is even more so in the data-centric market in which different applications (including voice) co-exist on different networks – 2G, 3G, 4G, Wi-Fi, Fibre, Copper, etc.

Telco 2.0 Happy Pipers also need to expand their addressable market in order to thrive – into Infrastructure Services, M2M, Embedded Connectivity and, in some cases, into Enterprise ICT including bespoke vertical industry solutions.  For sure this requires some new Service Development capabilities but, perhaps more importantly, also new partnerships – both in terms of service development and delivery – and a greater focus on Customer Experience Management and ‘Customer data/Big data’ in order to deliver valuable solutions to demanding enterprise customers.

For a Telco 2.0 Service Provider, the range of new capabilities required is even greater:

  • The ability to develop new platform and end-user (consumer and enterprise) services.
  • Brand management – not just creating a stolid telecoms brand but a vibrant end-user one.
  • New partners in other industries – financial services, media, advertising, start-ups, developers and so forth.


Figure 1: Capabilities needed for different Telco 2.0 strategies

Fig1 Capabilities need for different Telco 2.0 Strategies

Source: STL Partners/Telco 2.0

Most leading CSPs are pursuing a Telco 2.0 ‘Service Provider’ strategy

STL Partners analysis suggests that the majority of CSPs (and certainly all the tier 1 and 2 players) have at least some aspirations as a Telco 2.0 Service Provider.  Several, such as AT&T, Deutsche Telekom Orange, SingTel, Telefonica and Telenor, have been public with their ‘digital services’ aspirations.

But even more circumspect players such as Verizon and Vodafone which have to date largely focused on core telecommunications services have aspirations to move beyond this.  Verizon, for example, is participating in the ISIS joint venture on payments, albeit something of a slow burn at present.  Vodafone has also pushed into payments in developing markets via its successes with mPesa in Kenya and is (perhaps a slightly reluctant) partner in the WEVE JV in the UK on digital commerce.

Further back in their Telco 2.0 development owing to the attractiveness of their markets from a Telco 1.0 perspective are the players in the rapidly developing Middle Eastern and Asian markets such as Axiata, Etisalat, Mobily, Ooredoo, and Zain.  These players too aspire to achieve more than Happy Piper status and are already pushing into advertising, content and payments for consumers and M2M and Cloud for enterprises.

Telco 2.0 Service Providers are adopting different organisation designs

It is clear that there is no consensus among management about how to implement Telco 2.0 services. This is not surprising given how new it is for telecoms operators to develop and deliver new services – innovation is not something associated with telcos.  Everyone is learning how to take their first tentative steps into the wonderful but worrisome world of innovation – like toddlers stepping into the shallow beach waters of the ocean.

There is no tried and tested formula for setting up an organisation that delivers innovation but there is consensus (among STL Partners’ contacts at least) that a different organisation structure is needed to the one that manages the core infrastructure business.  Most also agree that the new skills, partnerships, operational and financial model associated with Telco 2.0 innovation needs to be ring-fenced and protected from its mature Telco 1.0 counterpart.

The degree of separation between the old and new is the key area of debate.  We lay out the broad options in Figure 2.

Fig 2 Organisation design models for Telco 2.0 Service Innovation

Fig 2 Organisation design models for Telco 2.0 Service Innovation

Source: STL Partners/Telco 2.0

For some, a central independent strategy unit that identifies potential innovations and undertakes an initial evaluation is a sufficient degree of separation.  AT&T and Verizon in the US have gone down this route – see Figure 3.

Fig 3 Organisation design approaches of 9 CSPs across 4 regions

Fig 3 Organisation design approaches of 9 CSPs across 4 regions

Source: STL Partners/Telco 2.0

In this model, ideas that are deemed promising are handed over the operating units to develop and deliver where, frankly, many are ignored or wallow in what one executive described to us as ‘Telco goo’ – the slow processes associated with the 20-year investment cycles of an infrastructure business.

Players such as Etisalat, Mobily and Ooredoo that are taking their first steps into Telco 2.0 services, but harbouring great aspirations, have gone a step further than this and set up Central Innovation Units.   In additional to innovation ideation and evaluation, these units typically undertake piloting, investment and, in some cases, some modest product development.  This approach is a sensible ‘first step’ into innovation and echoes the earlier attempts by many multi-national European players in the early 2000’s that had central group marketing functions that undertook proposition development for several countries.  The benefit is that the company can focus most resources on growth in existing Telco 1.0 services and Telco 2.0 solutions do not become a major distraction.  The downside is that Telco 2.0 services are seen as small and distant are always far less important than voice, messaging and connectivity services or devices ranges that can make a big impact in the next 3-6 months.

Finally, the most ambitious Telco 2.0 Service Providers – Deutsche Telekom, SingTel, Telenor, Telefonica and others – have developed separate New Business Units  The Telco 2.0 New Business Unit is given end-to-end responsibility for Telco 2.0 services.  The units find, develop, launch and manage new digital services and have full P&L responsibility.

STL Partners has long been a fan of this approach.  Innovation is given room to develop and grow under the guidance of senior management.  It has a high profile within the organisation but different targets, processes, people and partnerships to the core business which, left unchecked, would intentionally or unintentionally kill the new ‘rival’ off.

Five Principles for developing a Telco 2.0 New Business Unit

  1. Full control and responsibility.  The unit must have the independence from the core business to be able to control its own destiny and not be advertently or inadvertently impeded by the core business.  Telefonica, for example, went as far as to give its unit a separate physical location in central London.
  2. Senior management support.  While the unit is largely independent, it must be part of the corporate strategy and decisions about it must be made at the highest level.  In other words, the unit must be tied to the core business right at the top of the organisation – it is not completely free and decisions must be made for the overall good of the company.  Sometimes those decisions will be to the benefit or detriment of either the core business or the new business unit.  This is inevitable and not a cause for alarm – but these decisions need to be considered carefully and rationally by the senior team.
  3. Go OTT to start with.  One of the challenges faced by senior managers is how to leverage the capabilities of the core business – the network, customer data, retail outlets, brand, etc. – in the digital services offered by the new unit.  Clearly, it makes sense to use these assets to differentiate against the OTT players.  However, STL Partners recommends not trying to do this initially as the complexity of building successful interfaces between the new unit and the core business will prove too challenging.  Instead, establish some momentum with OTT services that the new unit can develop and deliver independently, without drawing on the core business, before then adding some specific core business capabilities such as location data, customer preference data or network QoS.
  4. Don’t forget to change management incentives …There is no point in filling the new business unit with senior management and fresh talent imbued with new skills and undertaking new business processes and practices unless they are clearly incentivised to make the right decisions!  It seems an obvious point but CSPs have a long and successful infrastructure legacy which means that management incentives are typically suitable for this type of business.  Managers typically have to hit high EBITDA margins, revenue targets that equate to around 50% of the capital base being generated a year, strong on-going capital investment – things that are at odds with a product innovation business (lower EBITDA margins, much lower capital intensity).  Management incentives need to change to reflect this and the fact that they business is a start-up not a bolt-on the core business.  These incentives need to be specific and can affect those in the core business as well as new unit.For example, if collaboration between the new unit and the core business units is a key requirement for long-term success (to build Telco 2.0 services that leverage core assets), then instigate a 360º feedback programme for all managers that measures how effectively they collaborate with their counter-parties in the other business units.  Scores here could be used to determine bonuses, share options or promotion – a sure way to instigate the required behaviour!
  5. …and investor metrics.  As mentioned above, a product innovation business has a different financial model to an infrastructure business.  Because of this, a new set of investor metrics is required focusing on lower margins and capital intensity.  Furthermore, users will often be a key metric rather than subscribers.  In other words, many users will not directly generate revenue (just as they do not for Google or Facebook) but remain an important driver of third-party sponsorship and advertising revenues.  Linked to this, ARPU will become a less important metric for the new business unit because the end user will be one of several revenue sources.

Many of the leading telecoms players have, therefore, done the right thing with the development of their digital units. So why have they struggled so much with culture clashes between the core telecoms business and the new digital innovations?  The answer lies in the way the units have been set up – their scope and role, the people that reside within them, and the processes and metrics that are used to develop and deliver services. This is covered in the next section of this report.

 

  • Even the boldest players are too Telco-centric with their digital business units
  • Defining traditional and new Telco 2.0 services
  • Current digital business units cover all the new Telco 2.0 services but should they?
  • Option: Reduce the scope of the Digital Business Units
  • Telefonica’s recent closure of Telefonica Digital
  • How might Telefonica’s innovation and ‘digital services’ strategy play out?

 

  • Figure 4: Defining Telco 2.0 new services
  • Figure 5: The mixed bag of services found in current digital business units
  • Figure 6: Separate new Telco 2.0 Services from traditional telecoms ones
  • Figure 8: The organisation structure at Telefonica
  • Figure 9: Telefonica’s strategic options for implementing ‘digital services’

Mobile Broadband 2.0: The Top Disruptive Innovations

Summary: Key trends, tactics, and technologies for mobile broadband networks and services that will influence mid-term revenue opportunities, cost structures and competitive threats. Includes consideration of LTE, network sharing, WiFi, next-gen IP (EPC), small cells, CDNs, policy control, business model enablers and more.(March 2012, Executive Briefing Service, Future of the Networks Stream).

Trends in European data usage

  Read in Full (Members only)  Buy a single user license online  To Subscribe click here

Below is an extract from this 44 page Telco 2.0 Report that can be downloaded in full in PDF format by members of the Telco 2.0 Executive Briefing service and Future Networks Stream here. Non-members can subscribe here, buy a Single User license for this report online here for £795 (+VAT for UK buyers), or for multi-user licenses or other enquiries, please email contact@telco2.net / call +44 (0) 207 247 5003. We’ll also be discussing our findings and more on Facebook at the Silicon Valley (27-28 March) and London (12-13 June) New Digital Economics Brainstorms.

To share this article easily, please click:



Introduction

Telco 2.0 has previously published a wide variety of documents and blog posts on mobile broadband topics – content delivery networks (CDNs), mobile CDNs, WiFi offloading, Public WiFi, network outsourcing (“‘Under-The-Floor’ (UTF) Players: threat or opportunity? ”) and so forth. Our conferences have featured speakers and panellists discussing operator data-plan pricing strategies, tablets, network policy and numerous other angles. We’ve also featured guest material such as Arete Research’s report LTE: Late, Tempting, and Elusive.

In our recent ‘Under the Floor (UTF) Players‘ Briefing we looked at strategies to deal with some of of the challenges facing operators’ resulting from market structure and outsourcing

Under The Floor (UTF) Players Telco 2.0

This Executive Briefing is intended to complement and extend those efforts, looking specifically at those technical and business trends which are truly “disruptive”, either immediately or in the medium-term future. In essence, the document can be thought of as a checklist for strategists – pointing out key technologies or trends around mobile broadband networks and services that will influence mid-term revenue opportunities and threats. Some of those checklist items are relatively well-known, others more obscure but nonetheless important. What this document doesn’t cover is more straightforward concepts around pricing, customer service, segmentation and so forth – all important to get right, but rarely disruptive in nature.

During 2012, Telco 2.0 will be rolling out a new MBB workshop concept, which will audit operators’ existing technology strategy and planning around mobile data services and infrastructure. This briefing document is a roundup of some of the critical issues we will be advising on, as well as our top-level thinking on the importance of each trend.

It starts by discussing some of the issues which determine the extent of any disruption:

  • Growth in mobile data usage – and whether the much-vaunted “tsunami” of traffic may be slowing down
  • The role of standardisation , and whether it is a facilitator or inhibitor of disruption
  • Whether the most important MBB disruptions are likely to be telco-driven, or will stem from other actors such as device suppliers, IT companies or Internet firms.

The report then drills into a few particular domains where technology is evolving, looking at some of the most interesting and far-reaching trends and innovations. These are split broadly between:

  • Network infrastructure evolution (radio and core)
  • Control and policy functions, and business-model enablers

It is not feasible for us to cover all these areas in huge depth in a briefing paper such as this. Some areas such as CDNs and LTE have already been subject to other Telco 2.0 analysis, and this will be linked to where appropriate. Instead, we have drilled down into certain aspects we feel are especially interesting, particularly where these are outside the mainstream of industry awareness and thinking – and tried to map technical evolution paths onto potential business model opportunities and threats.

This report cannot be truly exhaustive – it doesn’t look at the nitty-gritty of silicon components, or antenna design, for example. It also treads a fine line between technological accuracy and ease-of-understanding for the knowledgeable but business-focused reader. For more detail or clarification on any area, please get in touch with us – email mailto:contact@stlpartners.com or call +44 (0) 207 247 5003.

Telco-driven disruption vs. external trends

There are various potential sources of disruption for the mobile broadband marketplace:

  • New technologies and business models implemented by telcos, which increase revenues, decrease costs, improve performance or alter the competitive dynamics between service providers.
  • 3rd party developments that can either bolster or undermine the operators’ broadband strategies. This includes both direct MBB innovations (new uses of WiFi, for example), or bleed-over from adjacent related marketplaces such as device creation or content/application provision.
  • External, non-technology effects such as changing regulation, economic backdrop or consumer behaviour.

The majority of this report covers “official” telco-centric innovations – LTE networks, new forms of policy control and so on,

External disruptions to monitor

But the most dangerous form of innovation is that from third parties, which can undermine assumptions about the ways mobile broadband can be used, introducing new mechanisms for arbitrage, or somehow subvert operators’ pricing plans or network controls. 

In the voice communications world, there are often regulations in place to protect service providers – such as banning the use of “SIM boxes” to terminate calls and reduce interconnection payments. But in the data environment, it is far less obvious that many work-arounds can either be seen as illegal, or even outside the scope of fair-usage conditions. That said, we have already seen some attempts by telcos to manage these effects – such as charging extra for “tethering” on smartphones.

It is not really possible to predict all possible disruptions of this type – such is the nature of innovation. But by describing a few examples, market participants can gauge their level of awareness, as well as gain motivation for ongoing “scanning” of new developments.

Some of the areas being followed by Telco 2.0 include:

  • Connection-sharing. This is where users might link devices together locally, perhaps through WiFi or Bluetooth, and share multiple cellular data connections. This is essentially “multi-tethering” – for example, 3 smartphones discovering each other nearby, perhaps each with a different 3G/4G provider, and pooling their connections together for shared use. From the user’s point of view it could improve effective coverage and maximum/average throughput speed. But from the operators’ view it would break the link between user identity and subscription, and essentially offload traffic from poor-quality networks on to better ones.
  • SoftSIM or SIM-free wireless. Over the last five years, various attempts have been made to decouple mobile data connections from SIM-based authentication. In some ways this is not new – WiFi doesn’t need a SIM, while it’s optional for WiMAX, and CDMA devices have typically been “hard-coded” to just register on a specific operator network. But the GSM/UMTS/LTE world has always relied on subscriber identification through a physical card. At one level, it s very good – SIMs are distributed easily and have enabled a successful prepay ecosystem to evolve. They provide operator control points and the ability to host secure applications on the card itself. However, the need to obtain a physical card restricts business models, especially for transient/temporary use such as a “one day pass”. But the most dangerous potential change is a move to a “soft” SIM, embedded in the device software stack. Companies such as Apple have long dreamed of acting as a virtual network provider, brokering between user and multiple networks. There is even a patent for encouraging bidding per-call (or perhaps per data-connection) with telcos competing head to head on price/quality grounds. Telco 2.0 views this type of least-cost routing as a major potential risk for operators, especially for mobile data – although it also possible enables some new business models that have been difficult to achieve in the past.
  • Encryption. Various of the new business models and technology deployment intentions of operators, vendors and standards bodies are predicated on analysing data flows. Deep packet inspection (DPI) is expected to be used to identify applications or traffic types, enabling differential treatment in the network, or different charging models to be employed. Yet this is rendered largely useless (or at least severely limited) when various types of encryption are used. Various content and application types already secure data in this way – content DRM, BlackBerry traffic, corporate VPN connections and so on. But increasingly, we will see major Internet companies such as Apple, Google, Facebook and Microsoft using such techniques both for their own users’ security, but also because it hides precise indicators of usage from the network operators. If a future Android phone sends all its mobile data back via a VPN tunnel and breaks it out in Mountain View, California, operators will be unable to discern YouTube video from search of VoIP traffic. This is one of the reasons why application-based charging models – one- or two-sided – are difficult to implement.
  • Application evolution speed. One of the largest challenges for operators is the pace of change of mobile applications. The growing penetration of smartphones, appstores and ease of “viral” adoption of new services causes a fundamental problem – applications emerge and evolve on a month-by-month or even week-by-week basis. This is faster than any realistic internal telco processes for developing new pricing plans, or changing network policies. Worse, the nature of “applications” is itself changing, with the advent of HTML5 web-apps, and the ability to “mash up” multiple functions in one app “wrapper”. Is a YouTube video shared and embedded in a Facebook page a “video service”, or “social networking”?

It is also really important to recognise that certain procedures and technologies used in policy and traffic management will likely have some unanticipated side-effects. Users, devices and applications are likely to respond to controls that limit their actions, while other developments may result in “emergent behaviours” spontaneously. For instance, there is a risk that too-strict data caps might change usage models for smartphones and make users just connect to the network when absolutely necessary. This is likely to be at the same times and places when other users also feel it necessary, with the unfortunate implication that peaks of usage get “spikier” rather than being ironed-out.

There is no easy answer to addressing these type of external threats. Operator strategists and planners simply need to keep watch on emerging trends, and perhaps stress-test their assumptions and forecasts with market observers who keep tabs on such developments.

The mobile data explosion… or maybe not?

It is an undisputed fact that mobile data is growing exponentially around the world. Or is it?

A J-curve or an S-curve?

Telco 2.0 certainly thinks that growth in data usage is occurring, but is starting to see signs that the smooth curves that drive so many other decisions might not be so smooth – or so steep – after all. If this proves to be the case, it could be far more disruptive to operators and vendors than any of the individual technologies discussed later in the report. If operator strategists are not at least scenario-planning for lower data growth rates, they may find themselves in a very uncomfortable position in a year’s time.

In its most recent study of mobile operators’ traffic patterns, Ericsson concluded that Q2 2011 data growth was just 8% globally, quarter-on-quarter, a far cry from the 20%+ growths seen previously, and leaving a chart that looks distinctly like the beginning of an S-curve rather than a continued “hockey stick”. Given that the 8% includes a sizeable contribution from undoubted high-growth developing markets like China, it suggests that other markets are maturing quickly. (We are rather sceptical of Ericsson’s suggestion of seasonality in the data). Other data points come from O2 in the UK , which appears to have had essentially zero traffic growth for the past few quarters, or Vodafone which now cites European data traffic to be growing more slowly (19% year-on-year) than its data revenues (21%). Our view is that current global growth is c.60-70%, c.40% in mature markets and 100%+ in developing markets.

Figure 1 – Trends in European data usage

 Trends in European Data Usage
 

Now it is possible that various one-off factors are at play here – the shift from unlimited to tiered pricing plans, the stronger enforcement of “fair-use” plans and the removal of particularly egregious heavy users. Certainly, other operators are still reporting strong growth in traffic levels. We may see resumption in growth, for example if cellular-connected tablets start to be used widely for streaming video. 

But we should also consider the potential market disruption, if the picture is less straightforward than the famous exponential charts. Even if the chart looks like a 2-stage S, or a “kinked” exponential, the gap may have implications, like a short recession in the economy. Many of the technical and business model innovations in recent years have been responses to the expected continual upward spiral of demand – either controlling users’ access to network resources, pricing it more highly and with greater granularity, or building out extra capacity at a lower price. Even leaving aside the fact that raw, aggregated “traffic” levels are a poor indicator of cost or congestion, any interruption or slow-down of the growth will invalidate a lot of assumptions and plans.

Our view is that the scary forecasts of “explosions” and “tsunamis” have led virtually all parts of the industry to create solutions to the problem. We can probably list more than 20 approaches, most of them standalone “silos”.

Figure 2 – A plethora of mobile data traffic management solutions

A Plethora of Mobile Data Traffic Management Solutions

What seems to have happened is that at least 10 of those approaches have worked – caps/tiers, video optimisation, WiFi offload, network densification and optimisation, collaboration with application firms to create “network-friendly” software and so forth. Taken collectively, there is actually a risk that they have worked “too well”, to the extent that some previous forecasts have turned into “self-denying prophesies”.

There is also another common forecasting problem occurring – the assumption that later adopters of a technology will have similar behaviour to earlier users. In many markets we are now reaching 30-50% smartphone penetration. That means that all the most enthusiastic users are already connected, and we’re left with those that are (largely) ambivalent and probably quite light users of data. That will bring the averages down, even if each individual user is still increasing their consumption over time. But even that assumption may be flawed, as caps have made people concentrate much more on their usage, offloading to WiFi and restricting their data flows. There is also some evidence that the growing numbers of free WiFi points is also reducing laptop use of mobile data, which accounts for 70-80% of the total in some markets, while the much-hyped shift to tablets isn’t driving much extra mobile data as most are WiFi-only.

So has the industry over-reacted to the threat of a “capacity crunch”? What might be the implications?

The problem is that focusing on a single, narrow metric “GB of data across the network” ignores some important nuances and finer detail. From an economics standpoint, network costs tend to be driven by two main criteria:

  • Network coverage in terms of area or population
  • Network capacity at the busiest places/times

Coverage is (generally) therefore driven by factors other than data traffic volumes. Many cells have to be built and run anyway, irrespective of whether there’s actually much load – the operators all want to claim good footprints and may be subject to regulatory rollout requirements. Peak capacity in the most popular locations, however, is a different matter. That is where issues such as spectrum availability, cell site locations and the latest high-speed networks become much more important – and hence costs do indeed rise. However, it is far from obvious that the problems at those “busy hours” are always caused by “data hogs” rather than sheer numbers of people each using a small amount of data. (There is also another issue around signalling traffic, discussed later). 

Yes, there is a generally positive correlation between network-wide volume growth and costs, but it is far from perfect, and certainly not a direct causal relationship.

So let’s hypothesise briefly about what might occur if data traffic growth does tail off, at least in mature markets.

  • Delays to LTE rollout – if 3G networks are filling up less quickly than expected, the urgency of 4G deployment is reduced.
  • The focus of policy and pricing for mobile data may switch back to encouraging use rather than discouraging/controlling it. Capacity utilisation may become an important metric, given the high fixed costs and low marginal ones. Expect more loyalty-type schemes, plus various methods to drive more usage in quiet cells or off-peak times.
  • Regulators may start to take different views of traffic management or predicted spectrum requirements.
  • Prices for mobile data might start to fall again, after a period where we have seen them rise. Some operators might be tempted back to unlimited plans, for example if they offer “unlimited off-peak” or similar options.
  • Many of the more complex and commercially-risky approaches to tariffing mobile data might be deprioritised. For example, application-specific pricing involving packet-inspection and filtering might get pushed back down the agenda.
  • In some cases, we may even end up with overcapacity on cellular data networks – not to the degree we saw in fibre in 2001-2004, but there might still be an “overhang” in some places, especially if there are multiple 4G networks.
  • Steady growth of (say) 20-30% peak data per annum should be manageable with the current trends in price/performance improvement. It should be possible to deploy and run networks to meet that demand with reducing unit “production cost”, for example through use of small cells. That may reduce the pressure to fill the “revenue gap” on the infamous scissors-diagram chart.

Overall, it is still a little too early to declare shifting growth patterns for mobile data as a “disruption”. There is a lack of clarity on what is happening, especially in terms of responses to the new controls, pricing and management technologies put recently in place. But operators need to watch extremely closely what is going on – and plan for multiple scenarios.

Specific recommendations will depend on an individual operator’s circumstances – user base, market maturity, spectrum assets, competition and so on. But broadly, we see three scenarios and implications for operators:

  • “All hands on deck!”: Continued strong growth (perhaps with a small “blip”) which maintains the pressure on networks, threatens congestion, and drives the need for additional capacity, spectrum and capex.
    • Operators should continue with current multiple strategies for dealing with data traffic – acquiring new spectrum, upgrading backhaul, exploring massive capacity enhancement with small cells and examining a variety of offload and optimisation techniques. Where possible, they should explore two-sided models for charging and use advanced pricing, policy or segmentation techniques to rein in abusers and reward those customers and applications that are parsimonious with their data use. Vigorous lobbying activities will be needed, for gaining more spectrum, relaxing Net Neutrality rules and perhaps “taxing” content/Internet companies for traffic injected onto networks.
  • “Panic over”: Moderating and patchy growth, which settles to a manageable rate – comparable with the patterns seen in the fixed broadband marketplace
    • This will mean that operators can “relax” a little, with the respite in explosive growth meaning that the continued capex cycles should be more modest and predictable. Extension of today’s pricing and segmentation strategies should improve margins, with continued innovation in business models able to proceed without rush, and without risking confrontation with Internet/content companies over traffic management techniques. Focus can shift towards monetising customer insight, ensuring that LTE rollouts are strategic rather than tactical, and exploring new content and communications services that exploit the improving capabilities of the network.
  • “Hangover”: Growth flattens off rapidly, leaving operators with unused capacity and threatening brutal price competition between telcos.
    • This scenario could prove painful, reminiscent of early-2000s experience in the fixed-broadband marketplace. Wholesale business models could help generate incremental traffic and revenue, while the emphasis will be on fixed-cost minimisation. Some operators will scale back 4G rollouts until cost and maturity go past the tipping-point for outright replacement of 3G. Restrictive policies on bandwidth use will be lifted, as operators compete to give customers the fastest / most-open access to the Internet on mobile devices. Consolidation – and perhaps bankruptcies – may ensure as declining data prices may coincide with substitution of core voice and messaging business

To read the note in full, including the following analysis…

  • Introduction
  • Telco-driven disruption vs. external trends
  • External disruptions to monitor
  • The mobile data explosion… or maybe not?
  • A J-curve or an S-curve?
  • Evolving the mobile network
  • Overview
  • LTE
  • Network sharing, wholesale and outsourcing
  • WiFi
  • Next-gen IP core networks (EPC)
  • Femtocells / small cells / “cloud RANs”
  • HetNets
  • Advanced offload: LIPA, SIPTO & others
  • Peer-to-peer connectivity
  • Self optimising networks (SON)
  • M2M-specific broadband innovations
  • Policy, control & business model enablers
  • The internal politics of mobile broadband & policy
  • Two sided business-model enablement
  • Congestion exposure
  • Mobile video networking and CDNs
  • Controlling signalling traffic
  • Device intelligence
  • Analytics & QoE awareness
  • Conclusions & recommendations
  • Index

…and the following figures…

  • Figure 1 – Trends in European data usage
  • Figure 2 – A plethora of mobile data traffic management solutions
  • Figure 3 – Not all operator WiFi is “offload” – other use cases include “onload”
  • Figure 4 – Internal ‘power tensions’ over managing mobile broadband
  • Figure 5 – How a congestion API could work
  • Figure 6 – Relative Maturity of MBB Management Solutions
  • Figure 7 – Laptops generate traffic volume, smartphones create signalling load
  • Figure 8 – Measuring Quality of Experience
  • Figure 9 – Summary of disruptive network innovations

Members of the Telco 2.0 Executive Briefing Subscription Service and Future Networks Stream can download the full 44 page report in PDF format hereNon-Members, please subscribe here, buy a Single User license for this report online here for £795 (+VAT for UK buyers), or for multi-user licenses or other enquiries, please email contact@telco2.net / call +44 (0) 207 247 5003.

Organisations, geographies, people and products referenced: 3GPP, Aero2, Alcatel Lucent, AllJoyn, ALU, Amazon, Amdocs, Android, Apple, AT&T, ATIS, BBC, BlackBerry, Bridgewater, CarrierIQ, China, China Mobile, China Unicom, Clearwire, Conex, DoCoMo, Ericsson, Europe, EverythingEverywhere, Facebook, Femto Forum, FlashLinq, Free, Germany, Google, GSMA, H3G, Huawei, IETF, IMEI, IMSI, InterDigital, iPhones,Kenya, Kindle, Light Radio, LightSquared, Los Angeles, MBNL, Microsoft, Mobily, Netflix, NGMN, Norway, NSN, O2, WiFi, Openet, Qualcomm, Radisys, Russia, Saudi Arabia, SoftBank, Sony, Stoke, Telefonica, Telenor, Time Warner Cable, T-Mobile, UK, US, Verizon, Vita, Vodafone, WhatsApp, Yota, YouTube, ZTE.

Technologies and industry terms referenced: 2G, 3G, 4.5G, 4G, Adaptive bitrate streaming, ANDSF (Access Network Discovery and Selection Function), API, backhaul, Bluetooth, BSS, capacity crunch, capex, caps/tiers, CDMA, CDN, CDNs, Cloud RAN, content delivery networks (CDNs), Continuous Computing, Deep packet inspection (DPI), DPI, DRM, Encryption, Enhanced video, EPC, ePDG (Evolved Packet Data Gateway), Evolved Packet System, Femtocells, GGSN, GPS, GSM, Heterogeneous Network (HetNet), Heterogeneous Networks (HetNets), HLRs, hotspots, HSPA, HSS (Home Subscriber Server), HTML5, HTTP Live Streaming, IFOM (IP Flow Mobility and Seamless Offload), IMS, IPR, IPv4, IPv6, LIPA (Local IP Access), LTE, M2M, M2M network enhancements, metro-cells, MiFi, MIMO (multiple in, MME (Mobility Management Entity), mobile CDNs, mobile data, MOSAP, MSISDN, MVNAs (mobile virtual network aggregators)., MVNO, Net Neutrality, network outsourcing, Network sharing, Next-generation core networks, NFC, NodeBs, offload, OSS, outsourcing, P2P, Peer-to-peer connectivity, PGW (PDN Gateway), picocells, policy, Policy and Charging Rules Function (PCRF), Pre-cached video, pricing, Proximity networks, Public WiFi, QoE, QoS, RAN optimisation, RCS, remote radio heads, RFID, self-optimising network technology (SON), Self-optimising networks (SON), SGW (Serving Gateway), SIM-free wireless, single RANs, SIPTO (Selective IP Traffic Offload), SMS, SoftSIM, spectrum, super-femtos, Telco 2.0 Happy Pipe, Transparent optimisation, UMTS, ‘Under-The-Floor’ (UTF) Players, video optimisation, VoIP, VoLTE, VPN, White space, WiFi, WiFi Direct, WiFi offloading, WiMAX, WLAN.