We have written it so many times in recent years; and operators and vendors have said it so many times – if 5G deployment is to be cost-justified, beyond a few showcase networks, it must support a wide range of high value enterprise services. The consumer broadband model is dead as a source of growth or decent profitability. Like voice before it, it will continue to deliver cashflow, market share and other good things. But no operator can build out a new network without additional sources of revenue and growth.
It all seems so obvious, especially as LTE enhancements, including ‘Gigabit LTE’ (see separate item), can easily support changes in consumer demand for years to come. Yet many operators are talking in worryingly conventional ways – their CTOs may have grand industrial and IoT visions when addressing conferences, but when they discuss actual roll-out plans, it is usually in terms of enhancing mobile broadband first, then adding new services at some unspecified time in the future, with none of those use cases fully defined.
Waiting to deploy networks which are optimized for IoT and industrial applications is sensible. Enhanced LTE will satisfy consumers for years, even those wanting new augmented reality (AR) or virtual reality (VR) experiences. Many of those enhancements can be done in software, with no need for ripping and replacing base stations (except legacy ones, and new antennas may be required).
Operators would often do well to wait until the 5G New Radio is fully adapted and proven for low latency, high density machine-to-machine communications, for instance. They may also wait until network slicing, which will greatly increase their ability to support large numbers of enterprises and specialist service providers, is a commercial reality, with standards in place. And they should certainly wait until the economics of the new use cases are worked out.
While ‘keeping up with the Jonese’ or making a splash at a sports event may not be sufficient justification for moving on from 4G to 5G, that does not mean doing nothing. Operators need to work closely with industry players, regulators and city or transport authorities, to help work out those use cases, and understand exactly what will be required of a mobile network in the ‘everything connected’ world. China Unicom’s new ownership structure (see separate item) is one way that this world could go, with industries that want optimized connectivity sharing the cost of delivering that (as the web giants never have).
Ericsson’s new CTO, Erik Ekudden, had some refreshingly clear insights into all this when he talked to journalists and analysts last week, as Ericsson unveiled a paper about key technology trends in the 5G era. He was clear that the company will focus its efforts heavily on enterprise infrastructure and services, a mirror of Nokia’s strategy, as both companies struggle to recover from recent downturns in their core mobile broadband networks business.
When Ericsson’s new CEO, Börje Ekholm, took the helm early this year, he pulled back from many of his predecessor’s areas of expansion, such as media and cloud services, and reverted to the Swedish firm’s core businesses in network infrastructure and software, putting profits at the center of his objectives. Now the task is to push those core offerings to a wider array of customers, looking beyond telcos, or encouraging telcos to increase their investments in order to support new enterprise revenue streams.
Where Ericsson differs from Nokia or Huawei in this respect is that they want to serve enterprise customers directly. Indeed, Nokia has recently announced managed services platforms for enterprise and IoT organizations or service providers, which could potentially cut its traditional MNOs out of the loop in some deals. This is something Ericsson has done in the past, especially with direct contracts to deliver services to car companies like Volvo, but under the new regime, it says it will build new enterprise opportunities using only two channels – its telco customers and strategic partner Cisco.
Ulf Ewaldsson, in his new role as head of digital services, confirmed that strategy in an interview with Reuters last month. “We will focus on telco clients and networks exclusively for now,” he said.
Though this will avoid some conflicts of interest, it also risks putting all Ericsson’s eggs in the telco basket, even though this is a sector with declining capex and growth. Telco spending growth has been falling for several years and in some markets there have been declines in absolute terms in some years. Operators are looking to reduce capex by turning to software-driven architectures and even open source hardware. Ericsson itself believes the mobile infrastructure market will contract by between 2% and 6% this year.
Other verticals, by contrast, are becoming fully connected for the first time – not just relying on wireless networks for mobile voice and data, but looking to transform their processes by linking every object in their business chain to the internet and each other. That will drive investment. Nokia estimates its addressable market in adjacent sectors is worth about €18bn ($20.3bn) and will grow at a CAGR of 13% over the next five years. By contrast, the telco networks market is worth €113bn ($127bn) but will grow at a CAGR of just 1% in the five-year period.
So far, however, the Swedish firm has not made much impact beyond its core base, so it may feel that the cost and risk of casting a wide net are no longer justified – especially if it can start to leverage the Cisco relationship more effectively.
In late 2014, Ericsson set a target of generating between 20% and 25% of revenue from non-telcos by 2020, up from 10% in 2013, but it has not managed to increase these sales much from the 10% level and is no longer sharing targets in this area.
The challenge is worsened by the fact that Ericsson is not well placed to sell big ticket platforms to the one category of service provider which does have massive capex potential – the webscale giants. Analyst Richard Kramer from Arete Research said: “Ericsson simply lacks the products to sell to the likes of Google, Facebook, or Amazon, which are the biggest incremental spenders on infrastructure.” Those companies, with initiatives like Facebook’s Open Compute Project and Telco Infra Project, are not only leaving traditional vendors in the cold – developing platforms inhouse and turning to open source suppliers – but are even helping to lure MNOs away from Ericsson, Nokia and Huawei and towards commoditized, white box systems. By contrast, Nokia recently announced a ‘petabit-class’ chip to target the webscale space.
In the paper, Ekudden has identified five key trends in the technology industry, based on inputs from its customers, which could help the company leap ahead in supporting enterprise platforms, via telcos or Cisco. They are:
an adaptable technology base which combines software and hardware to increase efficiency and reduce cost;
machine intelligence to handle the petabytes of data networks are producing each day;
automation of networks, and making the “interaction between operations and network much more intuitive and smooth”;
end-to-end IoT security with “hardware routes of trust” in every IoT device;
and network slicing, particularly to support the IoT and its potential to transform industries.
Ekudden believes that changes in the operational side of networks, which are necessary to make digital transformation a reality rather than an overused cliché, have been taking place slowly and are starting to deliver tangible benefits. “The important thing which we are looking at is that it’s a timing issue,” he told Telecom.com. “It’s when the technology and the business storylines match up and a spark is created. That’s how we feel about these trends now.”
He added: “There is still a way to go on the automation side, but we are making some real progress. In many cases, the efficiencies being created, meaning cash and resource can be redeployed. It is changing the whole way customers are managing their businesses.” That needs to be followed by machine intelligence to make sense of all the data and drive innovation – and to enable another level of automation through AI-driven networks.
Perhaps only then, when the fully automated, software-driven, secure, intelligent and IoT-ready network platform is devised, will the move to the 5G radio make sense, and the new use cases be fully supportable. The radio will deliver many performance and efficiency benefits for these new digital models, but it will not enable a new way of working unless the underlying network architecture is in place first.
And if telcos devise that architecture without sufficient reference to the industries they wish to support in future, they will be stuck with networks which are still optimized primarily for consumer broadband. Ekudden is clear that industry collaboration is vital to expanding the business, for vendors and operators, in the 5G phase. That will be especially true for a vendor which wants to grow its enterprise mainly via telcos – if those operators do not understand the needs of industry, the vertical market players will bypass them and turn to managed services providers from the IT and cloud world, or perhaps Nokia.
“We have an opportunity to use new technologies in the 5G area for a different business from previous generations, which were predominantly for consumers,” he told LightReading. “But we need to foster collaboration and partnerships to make sure this opportunity becomes real.”
That refocusing on enterprise and industrial customers is not unique to the new CTO. A few weeks ago, Arun Bansal, head of Ericsson’s businesses in Europe and Latin America, said that 5G will not deliver much increase in operator spending unless new “industrial use cases” promise them (and the network suppliers) an improvement in revenue growth.
Bansal believes European operators are convinced that 5G needs to be driven by enterprise services, but are taking a long term view, waiting until the business cases are clear. This contrasts with some operators in North America and east Asia, which want to test 5G as early as possible, often as a fixed line substitute or a showcase event such as the Tokyo Olympics of 2020.
Nokia takes a contrasting path to Ericsson:
Nokia and Ericsson having copied and tailed one another for so many years, it is interesting to see them evolving into very different companies in the 5G era. Ericsson is sticking to its traditional customer base and model, and hoping to keep sales growing with new architectures that will help telcos grow their own businesses. Nokia wants to reduce its reliance on the telco market and go directly to higher growth sectors, tapping into their need to pursue IoT and digital transformation strategies.
Around the time of Mobile World Congress in February this year, Nokia really showed its hand, paying lip service to keeping the operator in its B2B loop, but in reality describing a platform in which that operator might be bypassed (especially when shared spectrum is in play), or play a minor bit-pipe role.
Nokia has announced two key platforms for vertical and IoT services, IMPACT and WING. IMPACT enables enterprises and service providers to manage their own IoT networks and services, while WING allows them to outsource that work to Nokia itself. Billing WING as a “worldwide IoT network grid”, the Finnish firm is making its most aggressive move to fulfil its promise to move into business ‘adjacencies’, to restart growth and offset tough pressures in its core network equipment business.
That entails becoming a company driven by software and services, and it is very clear that managed services will be the biggest link in the value chain in many IoT markets. Deploying, connecting, managing and securing huge numbers of connected devices, and handling all the data they produce, will be far outside the comfort zone of most enterprises, governments and even operators – even more so than running a mobile network, and even in that market, managed services have been on the rise.
Managed services already account for about one-third of Nokia’s revenues, but are targeted mainly at the traditional telco base, and in that market, the firm has been overshadowed by the might of Ericsson. Nokia now hopes to deploy and manage networks and services for its key verticals too. In IoT managed services, the market is immature and there is everything to play for.
There is also, however, a long line of companies chasing the opportunity, including IT giants like HPE; the network vendors; some of the big operators; and vertically specific bighitters like GE. So WING will need to make a big splash, to give Nokia a good spot on the starting grid.
Like most IoT ‘platforms’ emerging from established giants, much of WING is a tying-together of existing capabilities, including the IMPACT software itself, and managed services offerings. Nokia is pushing the multinational angle in particular, arguing that cross-border IoT networks and operators just don’t exist yet. So it will work with a host of network partners round the world – wired and wireless operators plus specialist vertical networks and potentially unlicensed systems – to offer a global connectivity grid.
Of course, its own technology will enable the “command center” at the heart of the grid, with customers being able to manage devices, billing and customer care on a global basis via IMPACT. Nokia says it will “offer a full service model including provisioning, operations, security, billing and dedicated enterprise customer services from key operations command centers”.
Other types of services which Nokia offers, or has spoken of developing, could then be layered on top of this – security (see inset), data management, analytics and so on.
Nokia will also offer a white label version of WING for operators, which can then add international support to services they provide in their own markets. However, the launch still sounds the latest in a series of warning bells, of the impending clash between network suppliers and their core customers, the telcos.
The telcos could have been a logical channel to market for Nokia’s software and services, but instead it is appealing directly to enterprises and governments, and not making carriers into equal partners, even when they are providing connectivity.
This is because Nokia and its rivals need to work towards a day when telecoms is just one vertical among many that they serve, rather than their entire business. Nokia has been adapting key offerings for the vertical markets it is particularly targeting – public safety, energy, transport, government, manufacturing, “technological extra-large enterprises” and webscale providers. These, it announced last year, will be the heart of its managed services and IoT offerings, within its increasingly autonomous software business – one of its great hopes for kickstarting healthier growth levels from 2017.
The first priority is to target the large enterprises. “Technological extra-large enterprises are the first set we are targeting and there is a dedicated channel set up to address those, which is delivering the sort of results we said it would,” Twist said in response to journalist questions at the launch. Nokia has made particular progress recently in utility/smart grid projects and in the mining sector. Large enterprise customers include Oklahoma Gas & Electric, Swissgrid, Washington Gas and Rio Tinto, and Nokia says it now has over 500 clients in the public sector, transportat and utility markets.
These type of organizations increasingly need to be connected and digital, linking up large numbers of items, such as machinery and vehicles, which have not needed mobile connectivity before. Unlike telcos, organizations such as utilities, manufacturers and banks are rarely expert in networks, and should be open to entrusting this to a third party.
This is especially true when it comes to mobile enterprises or wireless-centric IoT strategies – however virtualized these networks may become, they will always need actual radios and antennas, and radio skills remain precious and rarefied.