Digital infracos are emerging as cousins of towercos. Where the towerco focused on macro, ground based towers and coverage, digital infracos focus on street furniture, in-building solutions and Wi-Fi. The two asset classes are not mutually exclusive, and there is much commonality in terms of business models and value drivers. But how do we understand and evaluate the business of a digital infraco through the familiar lens of how we might understand and evaluate a towerco?
Defining ’digital infracos’
We recognise an emerging new class of digital infracos specialising in the provision of high quality, reliable cellular and Wi-Fi coverage in landmark buildings, private enterprises and public spaces, both indoors and in dense-urban outdoor environments.
Pureplay digital infracos own and operate the Wi-Fi, Distributed Antenna Systems (DAS) and/or small cell equipment used to provision connectivity, typically leasing capacity to multiple customers for a fixed monthly fee [1]. As such, pureplay digital infracos are distinct from Wi-Fi, DAS and small cell technology developers, who create and sell the equipment but don’t operate a service, and are also distinct from digital infrastructure systems integrators, which deploy and maintain equipment owned by an MNO or venue [2]. Similarly, we don’t consider a single tenant digital infrastructure deployment owned by a single MNO as a digital infraco, although such assets are often ripe for acquisition and upgrade by pureplay digital infracos who then add value by leasing up the infrastructure to multiple tenants under a neutral host model.
How do digital infracos make money?
In some models, particularly in emerging markets, digital infracos monetise MNOs. In others, such as the model pioneered in the UK by StrattoOpencell, now Freshwave, the venues pay. And we increasingly see hybrid models monetising both operators and venues.
The heart of the digital infraco business model remains the creation of long-term, recurring contracted revenues, with inflation-linked escalators, from multiple credit-worthy customers.
While this is remarkably similar to the macro tower company business model, some digital infracos make almost as much money selling analytics, managed [3] and Smart City services to venues, government and enterprises as they do from their multiple MNO tenants. We are starting to see digital infracos form agreements with enterprise IoT and VAS platforms to locate and operate their platforms on their networks. And some digital infracos even generate revenues directly from private consumers, such as passengers paying for premium Wi-Fi service at an airport. The sale of advertising and sponsorship may also generate significant revenue.
What are typical pricing and Master Lease Agreement terms?
Small cells and DAS are priced on a fixed monthly fee. There may be add-ons where there is a pole or wall mounting, for overhead cabling et cetera. Escalators typically kick in from year two, together with penalty-backed lock ins.
Wi-Fi is often monetised through a combination of fixed fees per access point or per venue per month and/or a variable fee based on GBs consumed, and VAS bundled. We have also seen examples of revenue share-based pricing of Wi-Fi.
A professional digital infraco would seek seven to ten or even fifteen-year contracts for small cell and DAS, typically shorter (three to five years) for Wi-Fi, reflecting the relative capital intensity of each technology. You will also often see shorter contracts agreed with enterprise customers, where corporate policy limits tenor to five years, although autorenewal is often possible.
A robust digital infraco contract will have carefully controlled restrictions on cancellation. Cancellation may be permitted on grounds of poor performance, although a premium digital infraco should not be unduly concerned about such clauses.
In general, revenues are amortised over the term of a contract, rather than collected as upfront fees, as this tends to drive the value of the digital infraco.
This response is based on a typical professional digital infraco ’ note that you may find some deviation from these norms in different geographical markets, or where technology developers or systems integrators are diversifying into owning and leasing digital infrastructure.
What are digital infracos’ primary costs?
Digital infracos’ primary costs are bandwidth, technology equipment and license fees, operations and maintenance costs. There costs are near constants, indeed they may decline over time. Lease costs and energy are often included under Wi-Fi, as leasing computer room space is offset against provision of service to the venue. The lease may be negotiated into the package therefore free of separate charge for small cell and DAS, but in some models the digital infraco pays to rent a computer room.
Energy is typically a pass through.
This all sounds very similar to macro towercos. What are the main differences of digital infracos?
Whereas macro towercos typically provision only passive infrastructure, digital infracos will often provision active infrastructure for Wi-Fi and DAS. This amplifies the technical competencies required by digital infracos, and increases technology risk.
In most markets, small cells remain owned by the MNOs, with the digital infraco partner providing only the passive infrastructure (poles, cabling, last mile fibre, power systems et cetera). Digital infracos in an increasing number of countries will be able to acquire, own, operate and lease up multi-tenant small cells as the regulation matures in recognition of the compelling economics of shared digital infrastructure, particularly for 5G.
What are the primary addressable markets for digital infracos?
Key addressable markets include medium to large scale venues including malls, public and corporate real estate, hotels, hospitals, educational institutions, sports and entertainment venues. Transportation is an important sector, including airports, railway stations and underground connectivity on urban metros. Smart city projects will expand the footprint of digital infracos into outdoor urban locations, adding IoT services to their potential revenue streams.
How penetrated is the addressable market for digital infracos?
The scale and fragmentation of the opportunity make it difficult to quantify, but in most markets only the larger landmark buildings (1mn sq ft+) are served by advanced in-building solutions (IBS), leaving a high volume of sites unserved in the ’middleprise’ layers: tier two 200,000-1mn sq ft, and tier three 25,000-200,000 sq ft segments. As such, the ’land grab’ for IBS has barely begun, and the market could be compared to the macro tower industry 20 years ago.
Looking at the opportunity through a different lens, and using India as an example, the country is forecast to require 10mn Wi-Fi hotspots by 2022, supplemented by a further 250,000 small cells per annum. When one considers that between them, India’s seven leading independent infracos have around 1,250 IBS DAS locations, 750 Wi-Fi locations, and less than 5,000 small cells, you can see just how much of the addressable market remains to be captured in India.
How do digital infracos address these opportunities and drive organic growth?
Go to market strategies vary from one digital infraco to the next.
Having a strong relationship with MNOs’ network planning and IBS teams is critical regardless of the business model.
Many digital infracos leverage a substantial partner network, for example an enterprise segment can generate a substantial pipeline in partnership with office movers and office fitters. Partnerships with real estate asset managers can reveal instances where tenancies are running out at landmark buildings, a time when cellular quality of service can be an important part of lease renewal negotiations.
Another critical factor in successfully selling indoor connectivity as a service is knowing the market. Successful digital infracos will often have a sales lead for each segment of the market they address, for example an experienced real estate salesperson to promote the service to commercial real estate developers, a person with substantial experience of selling into the public sector to sell to government entities.
While on the one hand you have a broad partner programme and lead sourcing initiative to target locations and venues, on the other hand, many small cell and DAS opportunities come direct from MNOs, who identify locations requiring densification, and issue co-ordinates much like the search rings issued for macro towers.
Once new locations have been contracted, incremental organic growth is driven by leasing up sites to multiple MNOs, while also enabling new alternate revenue streams such as managed services, advertising and sponsorship.
How are digital infracos enabling smart cities?
We have seen many examples of forward-thinking digital infracos forming Public Private Partnerships with State and Municipal Governments to enable smart city programmes, often securing exclusive access to street furniture (lamp posts, bus stops etc) and public land to deploy digital infrastructure, often in return for a modest revenue share or provision of services to the government partner (e.g. CCTV, traffic cameras, waste and parking sensors).
Are there inorganic growth opportunities for digital infracos?
Consolidation among digital infracos seeking economies of scale is a growing trend, as is the acquisition of digital infrastructure ’servicecos’ ripe for conversion into asset ownership and lease-up models.
A third opportunity for inorganic growth comes from the acquisition of digital infrastructure solutions built by MNOs. While many MNOs don’t have substantial portfolios of existing Wi-Fi, DAS and small cell networks that can be easily monetised and third party operated, most MNOs will have built a few such systems [4], for example such systems may be deployed for priority enterprise customers or in partnership with a stadium, mall or airport. Especially where the technology and network design accommodate only a single tenant, such deployments are often considered a cost centre and under-invested, leaving the assets ripe to be acquired, upgraded and leased up by new entrant neutral host pureplay digital infracos.
While there are many tens of thousands, probably over a hundred thousand ’captive’ in-building solutions ’stranded’ on MNO balance sheets, a growing majority of MNOs prefer an ’asset light’ strategy, so they are often open to hiving off these assets and leasing them back from a neutral host digital infraco.
However, acquiring legacy IBS from MNOs is not always as simple as cherry picking the best sites when a technology refresh is required. In many instances, even if the deployed technology lends itself to multi-tenancy (as DAS often does), the systems deployed by one MNO may not satisfy the technical specifications to support other MNOs’ spectrum.
What are the barriers to entry for competitors to digital infracos?
Ownership of venue relationships is critical to protecting a digital infraco’s investment in indoor coverage. Outdoors, municipality / city relationships, especially when crystalised in a formal smart city programme, represent significant barriers to entry, particularly if they come with exclusive usage of street furniture (lamp posts, bus stops et cetera).
Where digital infracos own or control last mile fibre, it is unlikely that a competitor would overbuild to install a competitive network.
How should I undertake a meaningful comparison of the unit economics of macro towers and digital infrastructure such as IBS, DAS and small cells?
When considering indoor digital infrastructure, it is simplest to consider each location, rather than each individual access point, as comparable with a macro site.
When considering outdoor digital infrastructure, such as outdoor small cells, every ten small cells is typically comparable with one macro tower.
How should we measure the performance and value of a digital infraco?
Beyond traditional measures of book value, revenue, EBITDA, and discounted cash flow, digital infracos’ performance is best measured in terms of lease-up rate, node cash flow or the build multiple.
The lease-up rate, also known as the tenancy ratio, is simply the total number of tenants divided by the total number of sites. Much like macro towers, if a digital infraco can lease up their sites to an average of two or more tenants, creating a lease-up rate of two, it should generate healthy returns.
Node cash flow is the equivalent of tower cash flow, and is effectively a measure of gross profit: revenue less cost of sales, or in this case, revenue from leasing access points and value-added services, minus the direct cost of the nodes and the fibre (and excluding pass through costs such as energy).
The build multiple expresses revenue as a proportion of the cost to build the assets. So if you have a network which generates 2,000 units of currency per year and cost 10,000 units to build, your build multiple is five. If you have a ten-year contract, that’s healthy.
How do the contract structures of digital infracos compare to those of towercos?
To answer this question, it is best to focus on best practice contracts, which for both the towerco and digital infraco markets are typically characterised by long tenor, escalators and carefully controlled cancellation terms.
Some less formalised segments of the emerging digital infraco landscape are characterised by short contract tenor ’ often found where connectivity is provided by a serviceco rather than a pureplay digital infraco. In such circumstances, there is often an opportunity to extend contracts, thereby enhancing investibility.
Prospective investors in digital infracos should take comfort from the similarities between good digital infraco contracts, and good towerco contracts, given that the towerco business model is widely respected as generating robust returns.
What are the prospects for contract renewal?
We have seen the best digital infracos maintain renewal rates in excess of 95%.
Whoever controls the venue relationship effectively controls the contract. Venue managers are disinclined to allow their walls and ceilings to be opened up to re-cable new solutions, and equally reluctant to provide O&M access to multiple parties.
Meanwhile, MNOs continue to migrate toward an ’asset light’ business model ’ there is an increasing preference to move indoor and dense-urban connectivity from capex to opex, thus MNOs typically extend contracts, assuming they are satisfied with quality of service.
This adds up to a strong recipe for contract renewal.
Why don’t MNOs self-deploy their own digital infrastructure solutions?
Many MNOs have deployed their own indoor connectivity solutions in the past. However, such solutions are seldom cost effective when used by a single MNO, and competitive MNOs are often distrustful to share digital infrastructure.
As MNOs increasingly pursue ’asset light’ strategies, there is pent up demand to partner with neutral host digital infracos.
Why wouldn’t venues deploy their own connectivity solutions ’ why use a third party?
Connectivity is not a venue’s core competency, such that the vast majority of venues prefer a third party to deploy and maintain the technologies, and to negotiate terms with multiple MNOs. Digital infracos offer a low cost, plug-and-play solution, on a neutral infrastructure basis supporting all MNOs.
Who provides the fibre backhaul for digital infracos’ indoor and dense-urban outdoor connectivity solutions?
Fibre backhaul could be provided by any of three alternate parties: digital infracos either use their own fibre, or they use their MNO customer’s fibre, or dark fibre. Each has its own merits.
Where the digital infraco buys bandwidth from the MNO, this creates a symbiotic relationship between the digital infraco and MNO as the MNO may recover upwards of 50% of their fees from the backhaul fee. However, the digital infraco is typically free to go to market to find the cheapest source for backhaul, which may lead them to use dark fibre, where available.
Additional economics can be unlocked in cases where the digital infraco deploys/owns the fibre ’ by which we mean the digital infraco owning the fibre in the building ’ it’s fine for a dark fibre or mobile network operator to bring fibre to the front door. There is obviously both value and stickiness if the digital infraco can secure exclusive rights of way for fibre in the building.
While there are seldom regulatory restrictions protecting the digital infraco from being disintermediated by a fibre owner, in reality venues are strongly disinclined to re-deploy access points and the associated cabling.
Are tower companies that specialise in macro cell sites moving into the DAS and small cells space?
Some of the more forward-thinking towercos are adding DAS and small cells to their catalogue of connectivity solutions. However, pioneers such as Crown Castle and Digital Colony notwithstanding, most towerco DAS and small cell business units are dwarfed by their macro core businesses.
Many towercos prefer to buy rather than build their capabilities in DAS and small cells, exemplified by leading European towerco Cellnex’s acquisition of DAS pioneer CommsCon, and Indonesian market leading towerco Protelindo acquiring iForte.
How are digital infracos licensed and regulated?
Where a licensing and regulatory regime has been created for towercos, the same framework will often apply to digital infracos’ DAS and small cell offerings. However, many countries still have no formal regulations specifically for telecom infrastructure providers.
Wi-Fi services typically fall under ISP license regimes.
What could be the impact of the novel Coronavirus epidemic on digital infracos?
While there will inevitably be pros will and cons, the net outlook is positive for digital infracos.
CTOs reported a year’s worth of growth as mobile data usage leapt by 15-30% during lockdown, much of which is being sustained when countries re-open.
Governments have never been more acutely aware of the criticality of mobile telephony, which could have an important role to play in track and trace strategies. This will accelerate the progress of digitisation and smart city programmes, and may make rights of way, taxation and licensing regimes less burdensome in the medium term.
Some segments of digital infracos’ addressable market will be more affected than others. While the pandemic is having an adverse impact on footfall through hospitality, transport and retail venues, robust digital infrastructure will become increasingly critical as venues create ’touchless’ experiences and robotic human temperature screening. Other segments are more immediately positively affected, for example, digital infracos also provide connectivity for hospitals and other healthcare providers, where connectivity, telemedicine, automation and touchless experiences are becoming critical.
While the Work From Home trend may see central business districts less densely populated, a bigger trend will be Work From Anywhere ’ the ability to video conference anytime, anywhere, combined with a fresh impetus for co-sharing closer to where people live to minimise travel.
Please explain digital infracos’ enabling technologies.
Wi-Fi
Wi-Fi is a family of wireless network protocols used to create a local area network of devices and Internet access. Given that Wi-Fi is best suited to line-of-site use, a Wi-Fi network is enabled by multiple access points with a range between approximately 20m indoors and up to 150m outdoors.
Small cells
Small cells are lower powered, shorter range cells than macro cells, although the range of a small cell can vary between 10m indoors to 1-2km in outdoor rural settings. While small cells have less capacity than macro cells, they are critical to network densification ’ adding extra cells to areas where there is high demand for data, both indoors and outdoors. Types of small cells include femtocells, picocells and microcells, but they are distinct from Distributed Antenna Systems (DAS).
Distributed antenna system (DAS)
DAS are used to provide high quality cellular coverage in an indoor or outdoor space by installing a network of small antennas to serve as repeaters. These antennas are connected to a central control system which is in turn connected to the MNO’s base station. There are two main types of DAS: passive systems, which take signal from antennas on the roof of a building and disseminate the signal through leaky feeder cables; and active DAS, which uses fibre cables to transport the signal from rooftop antennas. Installation and equipment costs are relatively high for DAS, hence many DAS are shared by multiple tenants.
M-CASS
Multi Access Connectivity as a Shared Service, or M-CASS, is a technology agnostic platform able to integrate all connectivity solutions, including Wi-Fi, small cells and DAS, ready for 5G.
When analysing the technologies deployed by a digital infraco, one must look beyond vendor hyperbole to consider the market need one is trying to meet, and the go to market strategy. How is the digital infraco packaging the service? Are the technologies approved by all the MNOs in the market? Is approval blanket, or do you have to gain painstaking approvals, site by site? Critically, what is the management platform and control system, and do they make the systems scalable?
How does an MNO determine which of these technologies they might want to choose at a given location?
The selection of technology is fundamentally driven by traffic volume and value, and by the total cost of ownership horizon considered. For example, high value enterprise customers generate sufficient revenue to justify the relatively capitally intensive, longer term payback choice of a DAS, while small cells might be utilised to offload a congested network serving a busy street market. Wi-Fi is a go-to solution where relief of spectrum congestion is a priority, and the model can be adapted to economically serve relatively low ARPU environments.
As the use cases for small cells, DAS and Wi-Fi overlap and evolve, we will see migration from Wi-Fi to small cell and DAS; moving between solutions as customer needs change.
Which of these technologies is growing fastest and offers the best margins?
When considering the margin as a percentage of net service revenue, Wi-Fi has the lowest margin of the three technologies. DAS has the best margin but the highest capex cost (expensive equipment, plus complex civil works), and small cell somewhere in the middle ’ although small cell is anticipated to have the highest growth.
The differing margins reflect the relative capital intensity, differing revenue and tenancy profiles.
What is driving the forecast exponential growth of small cells, and what does this imply about the opportunity for digital infracos?
As coverage-driven macro tower market build outs plateau, with network planners and towercos running out of potential new urban macro sites to which to offload crowded networks, the search for capacity and latency shifts focus to small cells. Small cells are seen as an ideal solution to serve dense pockets of demand and new use cases both indoors and outdoors.
Pent up demand for small cells aggregates to several hundred thousand units worldwide over the coming two to three years but, outside the U.S., there are no neutral host players of scale as the opportunity is too new. The majority of MNOs are committed to partnering with third parties to share the cost of small cell deployments and bring down capex.
The unit economics of small cells compare favourably with those of macro towers. Even with a single tenant, annual fees typically deliver payback on the initial capex investment in small cells in less than two years, representing an IRR better than 30%. With many direct costs often passed through (lease costs and energy), incremental tenancies incur low incremental indirect costs, particularly at high volumes, fattening margins.
Is it possible to share one small cell between multiple tenants?
Multi-tenant, or multi-operator, small cell technology is on the market, including from vendors such as Ericsson (’Dot’), Huawei (’LampSite’) and CommScope. However, as sharing small cells is prohibited by regulations in many countries, most small cell deployments still consist of separate ’single tenant’ small cells for each MNO. In the U.S., for example, zoning restrictions currently mean each piece of ’street furniture’ can only accommodate a single tenant, meaning you need four lamp posts for four tenants, whereas in the U.K. multiple tenants can be located on the same structure.
As regulations mature in recognition of the criticality of shared infrastructure, particularly for 5G, multi-tenant small cells deployed by neutral host digital infracos are likely to become the norm, and ’co-location’ of a second tenant will be achieved digitally rather than through a physical site visit and installation of new equipment, thus accelerating time to market and improving margins for digital infracos.
Why is Wi-Fi still used to provision indoor connectivity?
While there is certainly merit to indoor cellular connectivity, in some markets the cost and scarcity of spectrum is such a problem that mobile data offload (MDO) becomes a priority for MNOs desperate for capacity to improve quality of service. MDO is a particular priority in emerging markets which lack legacy fixed line infrastructure. In India, for example, voice over-Wi-Fi has become incredibly popular.
The aforementioned drive toward ’touchless’ experiences as societies adapt to the Coronoavirus will further increase demand for robust Wi-Fi in public and enterprise spaces, while Wi-Fi can also play a critical role in affordably enabling many 4G/5G use cases.
