Ever since it tried and failed to merge with bigger rival O2 in 2016, Three UK has been the problem child of the UK telecom industry. Some 15 years after it first bowled into the country raving about 3G technology, it remains a teenager in need of a growth spurt, with just 12% of UK mobile customers. Its stroppy insistence that it can enjoy a successful career as a mobile-only service provider irritates those who champion fixed-mobile "convergence" as the key to a long and prosperous life. Where does Three go next?

The answer is growing clearer. Following a drip feed of announcements this year about network deals, and the acquisition of spectrum to support new 5G services, Three UK has been strutting its improved physique and sounding newly ambitious. Dave Dyson, its CEO, now says he wants to "double" the size of the business. And with few opportunities available to Three in the UK's saturated mobile market, it is through an attack on his rivals' broadband fiefdoms that he expects to deliver much of this growth. (See Digging Deeper Into Three UK's Transformation Program.)

Figure 1: Big Spender Dave Dyson, Three's CEO, says his company has made a £2 billion commitment to 5G rollout.

That does not mean Three is giving up its aversion to last-mile fixed lines. Rather like Verizon Communications Inc. (NYSE: VZ) in the US market, its aim is to use 5G as a residential broadband alternative to fiber, cable and copper. But Three's plan is on a much grander scale. While Verizon and others see 5G as a fill-in where fixed lines are unavailable, and where building them would be too expensive, Three appears to have near-nationwide aspirations. (See Verizon to Launch Fixed 5G Service on Oct. 1.)

It is not starting entirely from scratch. Last year, Three spent £250 million ($322 million) to acquire a residential broadband business called UK Broadband, which has been trading under the Relish brand. Rather like Three, it neither owns nor operates last-mile fixed lines, and instead uses 4G technology to provide broadband services. Today, it has between 150 and 200 4G mobile sites across central London, where it serves around 20,000 customers. But it has never developed a presence in other parts of the country. "It would have been too expensive to build a national footprint," said Dyson. (See 3 to Bag 5G Spectrum With £300M UK Broadband Move.)

The economics look very different with 5G, however, thanks largely to a re-architecting of Three's entire mobile network. Of real significance is a deal with SSE Enterprise Telecoms, a UK provider of fiber-optic networks and data-center services. SSE is connecting 20 state-of-the-art data centers Three has built across the UK in a major expansion of its data center footprint, which previously featured just three facilities in England. It is also helping Three to run fiber connections between mobile masts and "unbundled" fiber exchanges owned by UK telecom incumbent BT Group plc (NYSE: BT; London: BTA), providing "backhaul" for high-speed services. SSE's initial contract covers 177 of these exchanges, but Three's ultimate target is for about 440. (See Eurobites: Three UK Preps for 5G With Backbone Deal.)

Figure 2: Footprint Expansion Three is building new data centers across the UK (major locations shown in blue), having previously had just three in England (shown in red).

Gabriel Brown, a principal analyst with Heavy Reading, thinks a deal of that magnitude would cover about 80% of the UK's population, including all the country's "top markets." As Three upgrades its mobile network to support 5G technology, it could more economically factor residential broadband into its plans. "We'll load sites with the best antenna technology and spectrum we've got and do as much as we can in one visit," said Dyson. "When we launch we'll have spare capacity for which the incremental cost is low, and that is an opportunity to start to roll out home broadband on a much bigger footprint than Relish has today."

Want to know more about 5G? Check out our dedicated 5G content channel here on
Light Reading.

The important thing about the Relish takeover is the additional spectrum it brought. Following the UK's initial 5G spectrum auction earlier this year, Three now claims to have more 5G spectrum than any other UK operator. Indeed, when it comes to the "mid-band" spectrum (between 3.4GHz and 3.8GHz) now freed up for 5G, its 144MHz dramatically exceeds what any rival holds. With this advantage, Three reckons it will be able to provide peak 5G speeds of about 3 Gbit/s. No rival will manage more than 2 Gbit/s, it boasts. (See UK's £1.4B '5G' auction looks bad for industry.)

Three's overhaul of its radio access network should also help. Having developed its 3G network with Finland's Nokia Corp. (NYSE: NOK), and used South Korea's Samsung Electronics Co. Ltd. (Korea: SEC) on 4G, the operator has turned to China's Huawei Technologies Co. Ltd. in the 5G era. By relying on a sophisticated new antenna technology called massive MIMO, it aims to boost coverage and capacity on the 5G network. Phil Sheppard, Three's director of network strategy, reckons massive MIMO will obviate the need for network "densification," allowing Three to overlay its mid-band network on the same "grid" used for 3G and 4G services. Three is also counting on an overall capacity gain of three to five times compared with 4G. (See Nokia, Samsung Miss Out as Three UK Gives 5G Job to Huawei.)

Next page: Tricky maneuver

Tricky maneuver
Advancing into fiber and cable territory with a 5G offer will be a difficult maneuver, nevertheless. Companies like BT, Virgin Media Inc. (Nasdaq: VMED) and Vodafone UK are now investing in all-fiber connections that promise speeds any mobile technology will struggle to match. Heavy Reading's Brown admits to feeling a "bit skeptical" about the move. The average Relish customer, he notes, consumes about 80 gigabytes of data each month, while usage in typical fixed-line households tends to be much heavier.

Another challenge with mobile, he says, is providing connectivity for all homes in a zone. "You will only get a certain proportion of users that are serviceable because of line-of-sight conditions," he says. "Traditionally, in broadband marketing, you don't want to just access a proportion of the region -- you want everyone in the postcode -- and that proposition may be a problem for them."

Figure 3: Spectrum Boost Three's spectrum situation before and after recent takeover activity and the UK's 5G auction.

Despite these reservations, he is impressed with Three's latest network strategy. "It is a top-to-bottom refresh after the O2 merger wasn't approved," he says, referring to the deal that was shot down by regulatory authorities on competition grounds. There might also be advantages in using mid-band spectrum for residential broadband services, rather than the extremely high-frequency "millimeter wave" bands that have grabbed all the attention in the US. "3.5GHz is the traditional fixed wireless band and how it works is well understood," he says. (See Telefónica Eyes Alternative Buyers for UK Biz – Report.)

There can be little doubt that line of sight and propagation are much bigger issues for millimeter wave spectrum, simply because signals do not travel as far or penetrate walls as effectively in these higher ranges. If Three can take full advantage of massive MIMO technology and its generous spectrum holdings, it may be able to overcome the challenges it faces on the capacity side. "It might not be a Rolls Royce service, but it will be pretty good," says Brown.

Want to know more about 4G LTE? Check out our dedicated 4G LTE content channel here on Light Reading.

Whether it is good enough for a doubling in the size of Three's business is a hard question to answer, partly because Dyson is deliberately vague about the nature of this growth. "Being twice as big, whether in terms of revenues or margins or customers, will make us more secure and sustainable as a business," is all he will say. There is also no timeframe attached to this plan.

Clearly, there is an opportunity to "upsell" broadband to some of Three's existing mobile customers. This "convergence" ploy is exactly what BT is now focused on after integrating EE, the mobile giant it bought in 2016, with the rest of its business. Three's pricing competitiveness, moreover, is one of its strengths, and could help to lure customers who think BT's broadband is too expensive.

Next page: Mobile squeeze

Mobile squeeze
Outside this broadband market, the 5G opportunity looks far less compelling. Three's mobile customer base has continued to grow, numbering 13 million subscriptions at the end of June -- when parent company CK Hutchison last reported results -- up from 12 million a year earlier (although just 10.1 million subscriptions were classed as "active," compared with 10 million in June 2017). But any 5G-fueled bid for mobile market share would be risky, acknowledges Dyson. "If we doubled the mobile customer base it would have to come from somewhere and that would result in retaliation or reaction from other operators, so we have to be smart," he said.

Three's boss also plays down the prospect of a 5G sales boost from existing mobile customers. "I cannot predict what others will do but we didn't charge 4G at a premium," he said. Demanding higher fees in the early days of 5G, when it is not widely available, would also be hard to justify, he reckons. "I would not want to put barriers in the way of customers getting a better experience."

Figure 4: Big Shifts Three claims its overhaul will lead to various performance and cosat improvements.

Another untapped market for Three could be the enterprise sector, but the operator has had little to say about this so far. With the expansion of its data center footprint, it should be able to reduce the signaling delay on 5G networks. That, in turn, may lead to business with companies demanding this "low latency" connectivity for new applications and technologies, such as self-driving cars, remote surgery and factory automation. But these services are unproven and unlikely to appear for many years. Given Three's focus on what Dyson calls "high-value millennials," addressing business customers would also force Three to invest in new sales and marketing capabilities, at the very least. (See 5G Still More Like Rocket Fuel Than a Mission to Mars.)

To some extent that may be happening as Three teams up with brands like Easyjet (a low-cost airline) and Superdrug (a high-street retailer of health and beauty products that also counts CK Hutchison as owner). But the strategy is still very much about appealing to consumers Three has not previously courted. "It means we can more rationally target different parts of the UK population without spreading the Three brand across multiple segments," said Dyson.

For all the latest news from the wireless networking and services sector, check out our dedicated Mobile content channel here on Light Reading.

In the meantime, shareholders will have to ready themselves for some short-term pressure as Three invests in new systems and networks while older ones remain operational. Despite a 7% increase in revenues, to more than £2.4 billion ($3.1 billion), Three's earnings (before interest, tax, depreciation and amortization) fell 2% last year, to £702 million ($903 million), because of its spending program. Capital expenditure soared from £352 million ($453 million) in 2016 to £459 million ($591 million), and a similar amount is expected this year before the figure drops to about £400 million ($515 million) thereafter.

"Profitability growth has stagnated a bit and that will continue as we go through this year and until we start to decommission the old legacy systems, but we've always been a strategic, long-term player," said Dyson. "This is fundamental not just to maintaining competitiveness but to becoming stronger."

— Iain Morris, International Editor, Light Reading