How Will 5G Small Cell Costs Compare to FTTH?

Nobody yet really knows how fixed wireless enabled by a 5G network will compare, in terms of deployment cost, with fiber to home costs, except to say virtually everyone expects that cost to be less than FTTH.

The issue is “how much lower, per potential passing,” those costs will prove to be. Many of the potential data points (fixed wireless to high rise buildings; mesh networks for business or consumers) are so different from potential ubiquitous 5G small cell deployments that those other examples are not so useful. Nor can the other deployments fully capture the costs of dealing with line of sight impediments, when small cells might be deployed very densely, perhaps on every other light pole.

What is really different about dense small cell networks is that, for the first time, the total cost of the infrastructure might be dominated by the cost of the trunking network, not the radio access network.

It might not be unreasonable to assume that a fiber-deep network of the sort Verizon is building, which might be called a “fiber to light pole” deployment, would be less than, or close to, the cost of a fiber to node architecture. According to Nokia estimates, the trunking network should cost less than half the cost of fiber to the home, and conceivably just a quarter of FTTH cost.

source: Nokia

One Time When Scale Apparently is a Negative

Scale matters in telecommunications and network-delivered content. But some forms of scale, namely the number of channels offered to consumers, seem to have negative impact. In other words, offering “too many” channels leads to much lower revenue per channel, as you would expect.

The relationship between market share and revenue per channel is less clear, if only because some smaller providers also offer the smallest channel lineups, while some larger suppliers offer the most channels.

source: Ampere Analysis

Service Providers	Subscribers at End of 1Q 2017
Cable Companies
Comcast	22,549,000
Charter	17,147,000
Altice*	3,500,000
Mediacom**	832,000
Cable ONE	307,187
Other major private company***	4,275,000
Total Top Cable	48,610,187
Satellite Services (DBS)
DIRECTV	21,012,000
DISH-DBS^	12,173,000
Total DBS	33,185,000
Phone Companies
Verizon FiOS	4,681,000
AT&T U-verse	4,048,000
Frontier^^	1,065,000
Total Top Phone	9,794,000
Internet-Delivered
Sling TV^	1,355,000
DIRECTV NOW^^^	375,000
Total Internet-Delivered	1,730,000
Total Top Providers	93,319,187

It also is possible to note that suppliers will smaller subscriber bases also tend to carry more channels, which, all things being equal, also lowers revenue per channel.

Nobody yet really knows how fixed wireless enabled by a 5G network will compare, in terms of deployment cost, with fiber to home costs, except to say virtually everyone expects that cost to be less than FTTH.

source: Leichtman Research

No 5G Business Model; It Has to be Invented

It has to be said: the business model for 5G will have to be created, it is not intuitive. Like earlier networks, most users eventually will migrate from older networks to 5G. The issue is how then to pay for the new network, if the base of human users is essentially the same as exists for 4G and older networks.

“Faster internet access” is the logical answer. And there the evidence is that higher usage, when usage is rated, leads to higher average spending. And that points to one potential problem. If usage is “flat rate for all you can eat,” higher usage does not increase revenues.

So a 5G network deployment adds capital investment cost, but possibly little incremental revenue from human users.

Many will note that is precisely why internet of things is so crucial: it represents the best hope for incremental new value and revenue from new use cases in enterprise verticals where the users are sensors, not humans.

Some are blunt. “Unlike 4G, there is no discernable use case that will encourage operators to roll out 5G networks,” says Juniper Research.

As a result, Juniper anticipates that increased investment from governmental bodies will be needed to encourage the development of these networks in most regions outside North America.

In many cases, that support might include measures such as spreading spectrum investment over some years.

source: Juniper Research

The 5G era also seems to be shifting the strategic function of fixed networks, adding value as backhaul for small cells expected to anchor much of the 5G network. Verizon has been most forthright in arguing that spectrum assets are only one way to increase capacity, something that always has been true.

In fact, some might well argue that, historically, shrinking cell radii has been a more-typical way of increasing capacity than adding new spectrum assets. In the 5G era, both will play important roles, especially in markets where regulators will release huge amounts of new spectrum for 5G and other networks.

Despite the huge increases in new spectrum assets (perhaps 10 times to 100 times as much as currently is available for all mobile purposes), some believe 5G spectrum will cost significantly less than did 4G spectrum.

5G is going to be different from 4G, says Juniper Research. Different enough, in fact, that business models, use cases and implementation will differ from 4G, indeed from all prior mobile generations, one might well argue.

There is a near certainty that brand-new revenue streams and use cases will disproportionately be created by enterprise customers supporting internet of things applications. That is not to say consumer internet access will not be important. Obviously, eventually most 4G users will migrate to using the 5G network. But it is unclear how much incremental revenue might be created, and for how long, by that substitution.

In the consumer internet access area, there is one important exception: use of the 5G platform to create fixed wireless connections with features that rival fiber-to-customer or hybrid fiber coax networks.

Juniper Research believes that use case will prove viable in both urban and rural areas. “The peak data rates of a 5G system will be higher than 10 Gbps but, more importantly, the cell-edge data rate (for 95 percent of users) should be 100 Mbps,” says Juniper Research. “This will allow the use of the mobile Internet as a reliable replacement for cable wherever needed.”

That will be more important in some markets. Lower-density markets will benefit, as those are areas where alternatives such as fiber to the premises are too costly.

In the consumer internet access area, it will likely prove to be the case that fixed wireless--allowing the mobile network to compete fully with the fixed network--will be the most-significant new use case.

Should Regulation be Technology Neutral?

Communications regulators always face issues when technology changes are rapid, the common complaint being that regulation lags technology change by quite some measure.

As the U.S. Federal Communications Commission says, “the fixed broadband industry continues to evolve.” For that reason, the FCC has issued a notice of inquiry on whether internet access should include both fixed and mobile delivery modes, and be tracked as a routine matter when assessing the state of internet access.

The proposal will raise hackles in some instances. Precisely at the point in time when we know gigabit and multi-gigabit internet access will be routinely supplied by mobile and fixed wireless networks operated by mobile service providers, some will say we should not “count” such access when assessing the state of the market.

In other words, the proposed FCC  “technology neutral approach” would continue to view access platforms in silos. That problem is not new. Even if there are important nuances, should legacy voice and VoIP not both be counted a methods for delivering voice services? And should cable modem services be counted separately from telco access services, since the platforms are different?

Are SONET-based data access solutions qualitatively different from Ethernet-based services, in terms of “counting” the usage of business data services?

Even if they use different technologies, are analog private branch exchange lines or trunks in service fundamentally different from IP equivalents?

Is digital TV something different from analog TV, or are both TV services, for purposes of measuring adoption and usage?

You get the point: technology change now is quite rapid in every communications and content delivery areas. And there are good reasons for using “technology neutral” methods of data collection and assessment.

Otherwise, voice, video or internet access delivered by satellite, cable TV, telco or independent ISPs are “different,” and not “parts of a single market.”

source: Nokia

Columbus Yellow Cab Using Verizon Vehicle Sharing Platform

Verizon is working with Columbus Yellow Cab to automate and streamline the cab reservation and payment process for its full fleet of cabs and drivers across Ohio. Columbus Yellow Cab is using Verizon’s Share solution that can track, manage and share fleet services and assets.

The vehicle sharing capability will help Yellow Cab drivers locate and rent available cabs, creating an on-demand, self-service approach to gaining access to vehicles.

Normally, Driver Partners lease vehicles from Yellow Cab on a daily, weekly or longer basis, forcing Driver Partners to repeat basic administrative duties and adding wait times to passenger pickups. Verizon’s Share solutions save Yellow Cab’s drivers time and money by enabling fleet sharing city-wide.

In addition to Yellow Cab, other companies working with Verizon’s Share Solutions platform include The Santa Clara Transportation Authority (VTA) and Swiftmile, the electronic bike (ebike) service; CruCar, a car sharing solution for airline crews, and A.A.S. Technologies, a security, networking and audio visual company.

Will 5G Drive Higher Business Customer Revenues?

So who is right: customers who mostly claim they are willing to spend more for 5G than 4G, or suppliers of 5G, who mostly believe a price premium will not be sustainable?

A recent global Gartner survey revealed that 75 percent of respondents at end-user organizations would be willing to pay more for 5G mobile capabilities. Only 24 percent of the survey's respondents would be unwilling to pay more for 5G than for 4G.

The survey also found that respondents from the telecom sector tend to see 5G migration as a matter of gradual and inevitable infrastructural change, rather than as an opportunity to generate new revenue, Gartner notes.

With the important caveat that survey respondents often say one thing and do another, the use cases suggest incremental service provider revenue is possible, but mostly because there are many new use cases related to internet of thing, meaning a new class of connections will be required.

Also, even if most of the end user respondents think their organizations would be prepared to pay more for 5G, few (eight percent) expect 5G to deliver cost savings or increase revenues. That might suggest the respondents are incorrect about the price increases.

End users tend to see 5G as a network evolution (59 percent) and only secondarily as an enabler of digital business (37 percent).

On the other hand, the end user responses also suggest why the belief in incremental revenues has some foundation. The survey found that almost half the respondents intend to use 5G to access videos and fixed wireless capabilities. The former is unlikely to drive incremental spending, but the latter could, in many instances.

The reason is akin to the possible implications of software-defined wide area networks. In principle, fixed wireless could allow new high-bandwidth connections where they are today not economically attractive, as SD-WAN might do the same for business data connections.

Some 57 percent of respondents believe that their organization’s main intention is to use 5G to drive Internet of Things (IoT) communication, and that might well prove to be correct.

Some might argue, with persuasiveness, that 5G will offer new capabilities primarily for very-high bandwidth applications, and conversely for very-low bandwidth connections used by some IoT sensors, or use cases where low latency is a necessity.

source: Gartner

Average Global Fixed Network Internet Access Price Close to $80/Month

On average, global fixed network internet access, adjusted on a purchasing power parity basis, runs between $50 a month to $80 a month, according to Point Topic. That might come as a surprise, since average prices near $80 hold for countries as disparate as India, Turkey and the United States.

As always with internet trends, there is a difference between peak, average and dynamic range of speeds. Some countries with relatively high “average” costs also have high dynamic range. In other words, a relatively small number of high cost connections can occur at the same time as large numbers of low-cost accounts.

In the second quarter of  2017, the average monthly charge for residential broadband services remained at $105, unchanged since the first quarter of 2017.  At the same time, the average download speed provided to residential subscribers continued to climb and was 135 Mbps, compared to 124 Mbps in the first quarter of 2017.

The average cost per megabit-per-second of speed continues to decline, Point Topic notes.  The average global cost-per-Mbps was US$0.78 at the end of the second quarter of 2017, compared to US$0.85 recorded at the end of the first quarter of  2017. As average speeds continue to climb, it would be reasonable to expect the average price “per Mbps” to continue to decline as well. That has been the trend for decades.

source: Point Topic

source: Point Topic

In some markets, such as the United States and Canada, the median price of internet access now has approached that of video subscription service. The difference is that profit margins for video are moderate to low, and declining, while profit margins for internet access are generally higher.

That is but one reason internet access has emerged as the anchor service for a fixed or mobile service provider. Among the other reasons are the high take rates for internet access and growing usage, while usage and average prices for other services decline.

The End of Cloud Computing?

Peter Levine sees another huge swing in computing architectures, away from cloud computing. The oscillation has been between centralized and distributed computing. The mainframe and cloud eras have been centralized; client-server was distributed.

With the coming era of mobile edge computing, Levine sees another historic change, back to distributed.

Competition Versus Investment: U.S. Regulators Might Face Dilemma European Regulators Have Faced

Telecom regulators and policymakers often face trade offs. They desire maximum feasible competition, but also maximum feasible investment in next generation facilities. The two objectives lie in a state of tension. Too much competition dampens investment. But too little competition also dampens investment.

The trick is finding the balance. In the European Community, regulators have fostered competition so well that incentives to invest now are viewed as having suffered. To create incentives for investment, rules on competition might need to be reworked. As it stands, many believe 5G investment in Europe will lag, compared to other developed markets.

Consolidation now seems inevitable in the U.S. telecom and content markets, but the U.S. market will not be alone. India’s mobile market seems to be on pace to reduce the number of national competitors roughly in half.

And Bell Labs predicts a massive consolidation of the world’s telecom service providers by about 2025.

Consolidation within any given segment of the market is designed to build scale, which boosts revenue and reduces costs. Part of the reason revenue gets a boost is that competition is reduced.

From an antitrust or competition review, it therefore matters how an agency defines “the market.” And we may be at one of those points where some amount of consolidation will be needed to create the cross-industry capabilities survivors will need.

That, in turn, means greater market power in one segment might be a necessary building block for survival in a new market that combines industry segments. In other words, less competition in one segment might be necessary to create the bundled or cross-segment capabilities survivors will require.

Consider the asymmetrical U.S. market positions. Verizon, AT&T, Comcast and Charter have fixed network assets expected to provide backhaul advantages when small cells underpin mobile network capabilities. Sprint and T-Mobile US do not have such assets, which arguably puts them at a disadvantage.

Verizon, AT&T and Comcast have content assets. Sprint and T-Mobile US do not have such assets. Dish Network owns spectrum, but no network and no content assets.

If you assume the future requires competitors that combine access assets (mobile and fixed) with “up the stack” revenue sources, then “less competition” in either fixed or mobile segments of the business, and fewer independent suppliers of content and apps, is going to happen.

So much will hinge on how antitrust officials define the relevant “market,” for purposes of determining competitive impact. By definition, consolidation reduces the number of competitors, and therefore has implications for competition.

But competition in telecom markets might also be heading for a period where ownership of several types of assets (fixed and mobile networks, content and application assets) is necessary for survival.

Much depends on how one views the coming market, and what winners will possess, in terms of assets and capabilities. That, in turn, will affect the definition of the “market,” and what consolidation means for competition in that market.

IoT Already is Driving Connection Growth

The attraction of 5G includes its projected importance as a driver of new internet of things connections. That trend (internet of things connections are a driver of new accounts) already has started.

Growth in the total number of mobile subscriptions was driven by a 14.2 percent increase in the number of machine-to-machine (M2M) connections in 2016, to 7.6 million in the U.K. market, says Ofcom.

source: Ofcom

That is crucial if one considers the chances 5G will drive significant additional revenue on the part of human users.

The problem is that consumers or businesses will only spend so much on all communications services. In the United Kingdom, for example, since 2011, household spending on fixed network voice and mobile voice has dropped.

Spending on fixed internet access services has grown, in large part because more consumers are paying incrementally more for faster speed services that cost more. Faster speed networks both are a response to higher usage, and arguably drive higher usage.

But there are limits to how much a human is willing to spend on access, or various communications and entertainment services. Generally speaking, consumers have substituted one product for another, even if there has arguably been growth in the monthly percent of household income spent on communications.

The point is that, even in the 5G, there are tough limits on how much incremental spending consumers are likely to entertain.

source: Ofcom

source: Ofcom

That is why internet of things and machine-to-machine connections are viewed as so significant: such connections represent a way to grow new accounts without pushing up against consumer spending resistance. Most IoT connections will be purchased by enterprises who have many other value drivers, against which the new connection cost is evaluated.

What are Prospects for Higher Mobile Data Spending on 5G Networks?

Here, in a nutshell, is the challenge for 5G networks, where it comes to the amount of incremental access revenue service providers can earn: consumers or businesses will only spend so much on all communications services. In the United Kingdom, for example, since 2011, household spending on fixed network voice and mobile voice has dropped.

Spending on fixed internet access services has grown, in large part because more consumers are paying incrementally more for faster speed services that cost more. Faster speed networks both are a response to higher usage, and arguably drive higher usage.

As a result, average revenue per gigabyte consumed is falling. “While average broadband revenue per connection has been increasing year on year, the average revenue per GB has been falling; from almost £1 per GB in 2011, to £0.15 in 2016 (a negative 5-year CAGR of 31%), Ofcom notes.

That is essentially the pattern seen in the capacity business, where volume increases do not keep linear pace with revenues earned supplying capacity. At the same time, big app providers increasingly build and operate their own networks, removing much of the demand from the potential capacity market.

source: Ofcom

source: Ofcom