One of the biggest challenges when designing a mobile network from scratch is how to access passive infrastructure that will support the new 5G architectures, and the dense projected footprint, at affordable cost.
Dish will need a wide variety of cell sites to support its schemes for a national network with sufficient capacity and density to differentiate it in terms of enterprise use cases and consumer quality of experience (QoE). Its new agreement with tower operator Crown Castle is a step in the right direction, though it is still unclear how much further Dish will need to go in terms of passive or active network sharing. Chairman Charlie Ergen said, in the 4G days, that building a network would not be viable without an infrastructure partner, but putative partnerships with Sprint or cable operators came to nothing.
The economics of 5G are different because Dish will deploy an open virtualized RAN on generic cloud hardware, but as noted in the introduction, much of the cost of a RAN is still embedded in physical elements from sites to fiber to edge cloud nodes (which need to be of very high quality to support demanding low latency RAN functions), and of course, active and passive antennas (to support an ever-widening range of 5G spectrum bands, including Massive MIMO for some higher frequency build-outs). On top of that is the need to improve in-building coverage and QoE, especially since Dish is increasingly focusing on the potential of B2B and wholesale enterprise models to sweeten its 5G business case.
For now, it gains nationwide coverage from its ‘enhanced MVNO’ agreement with T-Mobile – a condition of TMO’s merger with Sprint – but as it starts to identify the key target sectors and locations where it can deliver (and charge for) a premium service, it will need sites to be available wherever it needs them. That will not just mean towers, roofs and street furniture for radios, but a highly distributed edge cloud to support virtualized distributed units (vDUs), which in Dish’s plan for a dense open RAN will often sit very close to cell sites (see separate item below).
All this makes Crown Castle a strong partner since, of the major US towercos, it has been the most creative in expanding its portfolio beyond towers and roofs, and into small cells, in-building sites, fiber xHaul and edge cloud servers (although this has brought it into conflict with its activist investor Elliot Management, which wants it to offload some of these adjacent assets). Although Dish’s initial deal does not include small cells, there is every likelihood that it will be expanded in future, into the kind of multi-infrastructure contract that AT&T has pioneered with Crown and others.
Dish’s first deal with Crown Castle covers leasing on up to 20,000 towers over multiple years as well as fiber xHaul services. Dish also has the option to reduce its inhouse staff costs by using Crown’s pre-construction services, such as site engineering, zoning and permitting. Dish recently turned on its first 5G trial site and expects commercial network construction to start in the middle of next year.
No details were shared of the financial or other terms, but in a client note, Wall Street anlaysts at New Street Research said it seemed likely that Dish would prioritize usage of Crown sites where these are available in the desired location; which given the density of the towerco’s footprint, would give it a hefty share of the total business. Fellow analysts at Cowen & Co wrote that estimates from its channel checks indicated that Dish could pay a monthly fee of about $1,500 per tower per month, which would add up to $360m per year for Crown Castle across the 20,000 sites, excluding revenue from the fiber and network planning elements.
But revenues will not flow to Crown in earnest until full build-out starts in mid-2021 and it is likely to take Dish some years to require all 20,000 sites. However, the towerco does have the most urban-centric portfolio of the major US players, and the biggest investment in fiber (which it refers to as “horizontal towers”), and so its sites are well-suited to a 5G roll-out that will first emphasize areas of high capacity demand and high enterprise and consumer footprint.
New Street thinks there will be further deals to come, to extend Dish’s reach, and that eventually the business could be split evenly between the big three towercos (the others being American Tower and SBA), despite the clear headstart won by Crown. Analyst Spencer Kurn wrote: “CCI accounts for 30% of the US tower market, which could suggest that Dish intends to build to 65,000 towers if Dish’s network build is evenly distributed across the tower companies.”
However, that figure would be considerably more than the 50,000 macro and micro sites that Dish has been estimated to need. Other analysts, too, think Dish may previously have underestimated its infrastructure requirement. “This agreement suggests that the addressable market for Dish tower sites is larger than we previously estimated, likely +50,000 over time to reach nationwide coverage,” wrote analysts at Wells Fargo in an investor note. “We estimate that by FY2027, Dish could contribute $1bn-$1.2bn in annual tower revenues.”
The first phase of 5G build-out will require 15,000 sites to hit FCC coverage deadlines for June 2023, according to Dish’s chief commercial officer, Stephen Bye. He told the recent third quarter earnings call that Dish is on track for its target of covering 70% of the US population by mid-2023. That coverage target was agreed with the FCC as a compromise, in return for relaxing some of the original near-term mandates for population reach. That will give Dish slightly more time to configure its network, organize its supply chain, and wait for some key elements of its open platform to become more mature. It has chosen a number of vendors already, including Fujitsu for 5G radios, Nokia for the cloud-native 5G core, Mavenir and Altiostar for RAN network functions, VMware and others.
Its deal with Crown Castle, with the focus on fiber and services, and the likely expansion into small cells and edge servers, highlights the changing relationship of US mobile operators and towercos. A year ago, AT&T, which has been putting intense pressure on tower partners to reduce their prices, showed a more constructive side of the new towers deal when it announced its own multi-infrastructure agreement with American Tower; this, in turn, came more than a year after a novel, streamlined deal with Crown Castle, covering many kinds of sites and infrastructure in an efficient framework.
AT&T and American Tower announced a new, long term agreement in September 2019 and both claimed a victory in their multiyear rounds of negotiations, even though details of the terms, and the number of sites involved, were not revealed.
AT&T has been adding to its site partnership roster to create more price competition with the big towercos, and to increase its ability to access non-conventional sites in a standardized way (rather than having to negotiate for each lamp post or rooftop individually). It has signed deals directly with some cities for use of their infrastructure, though some of these have come at a hefty cost. It has also been working with alternative site providers such as Tillman, CitySwitch and Uniti Towers to build new towers, carrying lower rental rates, close to existing sites where AT&T believed it was paying excessive fees.
It seemed that, in return for additional business – presumably related to new sites, since we can assume per-site rentals have not risen in this deal – American Tower agreed to streamline its processes and charging structures, to reduce operational cost and paperwork. JR Wilson, AT&T’s VP of tower strategy and roaming, said in a statement that the deal would “drive mutual value and growth through a simplified leasing process designed to drive efficiency and flexibility improvements directly benefiting our speed in deploying the latest technologies”.
Crown Castle’s expanded leasing agreement with AT&T covers sites for outdoor small cells and for the FirstNet public safety network as well as macro sites. Currently, small cells are usually managed with completely separate site and operations processes to the main mobile network, and this has limited the ability to scale, especially if an MNO is having to deal with many small site owners rather than a single partner with a significant small site portfolio, as Crown Castle has been amassing, along with fiber assets for backhaul.
AT&T said the deal would streamline lease management and operations and give it improved flexibility to deploy whichever technologies were appropriate for particular locations and use cases. This will be vital to the economics of 5G, in which there will be far greater levels of density, and in which the lines between macro and small cells are already blurring.
Despite the activities of Elliot Management, Crown Castle has robustly defended the differentiation it gains from its focus on alternative infrastructure, and particularly on fiber. Speaking at an investor conference earlier this month, Crown’s CFO, Dan Schlanger, said fiber had been a differentiating factor in signing the deal with Dish.
“We put ourselves in a position of being a partner that extends beyond towers,” he said. “We think that was a big deal for them in trying to choose a partner. We believe having a large-scale tower operation focused in the top 100 markets in the USA, with fiber, was a differentiating factor for us and something that really skewed to our direction.”
In addition to 40,000 US. towers, Crown Castle has 80,000 route miles of fiber, as well as a small cell business with 70,000 sites either on air or under development. The Dish contract is an early proof point, Schlanger told the New Street conference, “that having that combination of solutions that we can provide helps us in winning business in the market”.