Introduction

It’s almost a cliché to talk about “the future of the network” in telecoms. We all know that broadband and network infrastructure is a never-ending continuum that evolves over time – its “future” is continually being invented and reinvented. We also all know that no two networks are identical, and that despite standardisation there are always specific differences, because countries, regulations, user-bases and legacies all vary widely.

But at the same time, the network clearly matters still – perhaps more than it has for the last two decades of rapid growth in telephony and SMS services, which are now dissipating rapidly in value. While there are certainly large swathes of the telecom sector benefiting from content provision, commerce and other “application-layer” activities, it is also true that the bulk of users’ perceived value is in connectivity to the Internet, IPTV and enterprise networks.

The big question is whether CSPs can continue to convert that perceived value from users into actual value for the bottom-line, given the costs and complexities involved in building and running networks. That is the paradox.

While the future will continue to feature a broader set of content/application revenue streams for telcos, it will also need to support not just more and faster data connections, but be able to cope with a set of new challenges and opportunities. Top of the list is support for “Connected Everything” – the so-called Internet of Things, smart homes, connected cars, mobile healthcare and so on. There is a significant chance that many of these will not involve connection via the “public Internet” and therefore there is a possibility for new forms of connectivity proposition evolving – faster- or lower-powered networks, or perhaps even the semi-mythical “QoS”, which if not paid for directly, could perhaps be integrated into compelling packages and data-service bundles. There is also the potential for “in-network” value to be added through SDN and NFV – for example, via distributed servers close to the edge of the network and “orchestrated” appropriately by the operator. But does this add more value than investing in more web/OTT-style applications and services, de-coupled from the network?

Again, this raises questions about technology, business models – and the practicalities of making it happen.

This plays directly into the concept of the revenue “hunger gap” we have analysed for the past two years – without ever-better (but more efficient) networks, the telecom industry is going to get further squeezed. While service innovation is utterly essential, it also seems to be slow-moving and patchy. The network part of telcos needs to run just to stand still. Consumers will adopt more and faster devices, better cameras and displays, and expect network performance to keep up with their 4K videos and real-time games, without paying more. Depending on the trajectory of regulatory change, we may also see more consolidation among parts of the service provider industry, more quad-play networks, more sharing and wholesale models.

We also see communications networks and applications permeating deeper into society and government. There is a sense among some policymakers that “telecoms is too important to leave up to the telcos”, with initiatives like Smart Cities and public-safety networks often becoming decoupled from the mainstream of service providers. There is an expectation that technology – and by extension, networks – will enable better economies, improved healthcare and education, safer and more efficient transport, mechanisms for combatting crime and climate change, and new industries and jobs, even as old ones become automated and robotised.

Figure 1 – New services are both network-integrated & independent

Source: STL Partners

And all of this generates yet more uncertainty, with yet more questions – some about the innovations needed to support these new visions, but also whether they can be brought to market profitably, given the starting-point we find ourselves at, with fragmented (yet growing) competition, regulatory uncertainty, political interference – and often, internal cultural barriers within the CSPs themselves. Can these be overcome?

A common theme from the section above is “Questions”. This document – and a forthcoming “sequel” – is intended to group, lay out and introduce the most important ones. Most observers just tend to focus on a few areas of uncertainty, but in setting up the next year or so of detailed research, Telco 2.0 wants to fully list and articulate all of the hottest issues. Only once they are collated, can we start to work out the priorities – and inter-dependencies.

Our belief is that all of the detailed questions on “Future Networks” can, it fact, be tied back to one of two broader, more over-reaching themes:

• What are the business cases and operational needs for future network investment?
• Which disruptions (technological or other) are expected in the future?

The business case theme is covered in this document. It combines future costs (spectrum, 4G/5G/fibre deployments, network-sharing, virtualisation, BSS/OSS transformation etc.) and revenues (data connectivity, content, network-integrated service offerings, new Telco 2.0-style services and so on). It also encompasses what is essential to make the evolution achievable, in terms of organisational and cultural transformation within telcos.

A separate Telco 2.0 document, to be published in coming weeks, will cover the various forthcoming disruptions. These are expected to include new network technologies that will ultimately coalesce to form 5G mobile and new low-power wireless, as well as FTTx and DOCSIS cable evolution. In addition, virtualisation in both NFV and SDN guises will be hugely transformative.

There is also a growing link between mobile and fixed domains, reflected in quad-play propositions, industry consolidation, and the growth of small-cells and WiFi with fixed-line backhaul. In addition, to support future service innovation, there need to be adequate platforms for both internal and external developers, as well as a meaningful strategy for voice/video which fits with both network and end-user trends. Beyond the technical, additional disruption will be delivered by regulatory change (for example on spectrum and neutrality), and also a reshaped vendor landscape.

The remainder of this report lays out the first five of the Top 10 most important questions for the Future Network. We can’t give definitive analyses, explanations or “answers” in a report of this length – and indeed, many of them are moving targets anyway. But taking a holistic approach to laying out each question properly – where it comes from, and what the “moving parts” are, we help to define the landscape. The objective is to help management teams apply those same filters to their own organisations, understand how can costs be controlled and revenues garnered, see where consolidation and regulatory change might help or hinder, and deal with users and governments’ increasing expectations.

The 10 Questions also lay the ground for our new Future Network research stream, forthcoming publications and comment/opinion.

Overview: what is the business case for Future Networks?

As later sections of both this document and the second in the series cover, there are various upcoming technical innovations in the networking pipeline. Numerous advanced radio technologies underpin 4.5G and 5G, there is ongoing work to improve fibre and DSL/cable broadband, virtualisation promises much greater flexibility in carrier infrastructure and service enablement, and so on. But all those advances are predicated on either (ideally) more revenues, or at least reduced costs to deploy and operate. All require economic justification for investment to occur.

This is at the core of the Future Networks dilemma for operators – what is the business case for ongoing investment? How can the executives, boards of directors and investors be assured of returns? We all know about the ongoing shift of business & society online, the moves towards smarter cities and national infrastructure, changes in entertainment and communication preferences and, of course, the Internet of Things – but how much benefit and value might accrue to CSPs? And is that value driven by network investments, or should telecom companies re-focus their investments and recruitment on software, content and the cloud?

This is not a straightforward question. There are many in the industry that assert that “the network is the key differentiator & source of value”, while others counter that it is a commodity and that “the real value is in the services”.

What is clear is that better/faster networks will be needed in any case, to achieve some of the lofty goals that are being suggested for the future. However, it is far from clear how much of the overall value-chain profit can be captured from just owning the basic machinery – recent years have shown a rapid de-coupling of network and service, apart from a few areas.

In the past, networks largely defined the services offered – most notably broadband access, phone calls and SMS, as well as cable TV and IPTV. But with the ubiquitous rise of Internet access and service platforms/gateways, an ever-increasing amount of service “logic” is located on the web, or in the cloud – not enshrined in the network itself. This is an important distinction – some services are abstracted and designed to be accessed from any network, while others are intimately linked to the infrastructure.

Over the last decade, the prevailing shift has been for network-independent services. In many ways “the web has won”. Potentially this trend may reverse in future though, as servers and virtualised, distributed cloud capabilities get pushed down into localised network elements. That, however, brings its own new complexities, uncertainties and challenges – it a brave (or foolhardy) telco CEO that would bet the company on new in-network service offers alone. We will also see API platforms expose network “capabilities” to the web/cloud – for example, W3C is working on standards to allow web developers to gain insights into network congestion, or users’ data-plans.

But currently, the trend is for broadband access and (most) services to be de-coupled. Nonetheless, some operators seem to have been able to make clever pricing, distribution and marketing decisions (supported by local market conditions and/or regulation) to enable bundles to be made desirable.

US operators, for example, have generally fared better than European CSPs, in what should have been comparably-mature markets. But was that due to a faster shift to 4G networks? Or other factors, such as European telecom fragmentation and sub-scale national markets, economic pressures, or perhaps a different legacy base? Did the broad European adoption of pre-paid (and often low-ARPU) mobile subscriptions make it harder to justify investments on the basis of future cashflows – or was it more about the early insistence that 2.6GHz was going to be the main “4G band”, with its limitations later coming back to bite people? It is hard to tease apart the technology issues from the commercial ones.

Similar differences apply in the fixed-broadband world. Why has adoption and typical speed varied so much? Why have some markets preferred cable to DSL? Why are fibre deployments patchy and very nation-specific? Is it about the technology involved – or the economy, topography, government policies, or the shape of the TV/broadcast sector?

Understanding these issues – and, once again, articulating the questions properly – is core to understanding the future for CSPs’ networks. We are in the middle of 4G rollout in most countries, with operators looking at the early requirements for 5G. SDN and NFV are looking important – but their exact purpose, value and timing still remain murky, despite the clear promises. Can fibre rollouts – FTTC or FTTH – still be justified in a world where TV/video spend is shifting away from linear programming and towards online services such as Netflix?

Given all these uncertainties, it may be that either network investments get slowed down – or else consolidation, government subsidy or other top-level initiatives are needed to stimulate them. On the other hand, it could be the case that reduced costs of capex and opex – perhaps through outsourcing, sharing or software-based platforms, or even open-source technology – make the numbers work out well, even for raw connectivity. Certainly, the last few years have seen rising expenditure by end-users on mobile broadband, even if it has also contributed to the erosion of legacy services such as telephony and SMS, by enabling more modern/cheaper rivals. We have also seen a shift to lower-cost network equipment and software suppliers, and an emphasis for “off the shelf” components, or open interfaces, to reduce lock-in and encourage competition.

The following sub-sections each frame a top-level, critical question relating to the business case for Future Networks:

• Will networks support genuinely new services & enablers/APIs, or just faster/more-granular Internet access?
• Speed, coverage, performance/QoS… what actually generates network value? And does this derive from customer satisfaction, new use-cases, or other sources?
• Does quad-play and fixed-mobile convergence win?
• Consolidation, network-sharing & wholesale: what changes?
• Telco organisation and culture: what needs to change to support future network investments?