FCC Moves to End Common Carrier Regulation of Internet Access

As expected, the U.S. Federal Communications Commission has voted to open a proceeding on internet access regulation, seeking what it calls a return to “light touch” regulation, and specifically, reconsidering regulation of internet access services under common carrier rules, something that was instituted in 2015. The proposal would return internet access to the “information service” category it had held for decades.

Some have supported the shift to common carrier regulation for a few stated reasons, including protecting edge providers (app providers) from “paid prioritization” schemes instituted by major internet service providers. Others opposed the common carrier rules because it limited permissible business models based on quality of service.

The ironic twist is that the viability of “paid prioritization” is questionable. Few seem to believe it is possible to create a sustainable market for such services.

Dr. George Ford, The Phoenix Center chief economist, argues that “no paid prioritization rule, from an economist’s point of view, should not be prohibited on a blanket basis.”

But that is not the same as arguing there is a clear consumer demand, or sustainable business case. “I’m not sure there is a business case.”

It is not that such possible payments are unusual. They are not. Two-sided markets in media, content and retailing are rather common, where an entity connects buyers and sellers, generating revenue from one set of partners, or sometimes from both. Any retailer--physical or online--can generate revenue from consumer buyers and product sellers.

Content networks earn money from distributors (satellite, cable TV, telco TV, Netflix, Prime) and advertisers. Distributors earn fees from subscribers (end users) as well as advertisers.

“Some sort of payment from a content provider is not that unusual,” says Russ Hanser, Wilkinson, Barker Knauer partner. “Is there a business model? I don’t know.”

There are analogies. The best example is the content delivery network (CDN), where content or app providers pay a third party to expedite delivery of packets over backbone networks, controlling latency. By extension, some have proposed there might be a market, or markets, related to providing latency protection issues in the consumer access services business.

The issue is whether direct revenue models can be created, and if so, how big those markets could be. The argument is that some apps, especially voice, videoconferencing and highly-transactional apps (gaming, for example) benefit enough from latency control that entities (app providers or end users) are willing to pay to have that capability. As always, consumers are resistant to value-added services, so value is less a question than propensity to pay.

Zero rating is sort of the mirror opposite of that issue. Where there clearly is a large market is allowing users to consume as much bandwidth as required to stream video or listen to audio. Where paid prioritization concerns revenue paid to ISPs by app providers, zero rating actually allows app providers to be used without any concern for mobile data usage limits.

In such cases, any revenue benefit is indirect, taking the form of higher new account additions or reduced churn. Eventually, ISP owners of their own subscription content services obviously will benefit more directly in the form of revenue earned by selling content subscriptions.

So far, it is not clear whether there is a market for paid prioritization. And, if so, it is not clear that the market is big.

Voice Hasn't Been Core for Mobile Since 3G

Looking at all the standards activity around 5G, it is easy enough to see reflections of the reality that voice is a fundamental feature for a communications provider, but drives a declining portion of revenue. All of the future revenue upside, in other words, comes from internet access and the services that access enables.  

Simply put, all the 5G activity is around mobile and fixed data. It is not clear that voice ever again will be a primary problem to be solved.

Some might argue that voice might become a bigger concern for mobile generation seven. Others might argue 3G was the last mobile network platform centrally concerned with voice.

Many speculate that mobile generation six could focus more on integrating mobile and satellite links, as 5G integrates licensed and unlicensed spectrum. That might also be an indication that mobile issues, as such, largely have been developed to a point where commercial attention turns elsewhere.

For the moment, most service providers in developed markets are trying to harvest voice revenues. In developing markets, voice remains a key revenue driver, particularly because subscription growth remains relatively high.

source: GSMA

In fact, we might be entering an era where mobile data becomes more akin to a feature than a revenue driver as well, following the pattern of voice services. To be sure, voice over LTE is expected to grow, in the 4G realm. Other forms of voice using an over-the-top method also will grow, though.

If you take the long view, some “news” items are not surprising. If you observe that mobile data revenues are peaking, then it is not too surprising when a decline occurs. Analyst Chetan Sharma notes that U.S. mobile data services revenue had seen quarter-over-quarter growth for 17 straight years until the first quarter of 2017.

In that quarter, revenue growth went negative.

Verizon saw its first-ever decline in service revenues, year over year.

For the first time, industry postpaid net-adds also were negative, while, for the first time, cars accounted for 50 percent of the total net-adds for the quarter, Sharma notes.

A product lifecycle top has been coming for some time. In developed markets, after “everyone who wants to use mobile” is doing so, and when “everyone who wants to use mobile data” is doing so, and when “everyone uses as much as they need”, the market simply is saturated.

And that means, as executives have understood, drives the search for big sources of new revenue to displace the lift once provided by mobile data services.

So 5G is not simply the latest in a series of mobile network generations, defined in part by air interface or data rate. In fact, that often is the way 5G is described.  No, 5G is something else, a deliberate attempt to build a network with intrinsic support for non-human users (internet of things).

Some might note that this is precisely what many have called pervasive computing, a world in which almost anything can have computing and communications embedded.

That might not be so obvious at first, as the first “at scale” deployments of 5G will focus on extending services for humans (“enhanced mobile broadband”). Ironically, though mobile data is reaching saturation, the first “at scale” revenue benefits and value might well be generated by mobile or fixed broadband.

Most industry executives seem to agree that, at first, most of the customer-observed value will come from higher data rates. None of that should obscure the more fundamental hope, though. And that is that 5G networks will enable a big new internet of things business delivering services to non-human users.

Among the biggest hoped-for changes is in business model. Industry executives believe that new revenues disproportionately will come from enterprises, not consumers, undoubtedly a reflection of the expectation that the incremental new revenue streams will be created by various internet of things services and applications.

5G Really Will be Quite Different

It is easy to argue that 5G is different from prior mobile network generations. It is the first network intentionally designed to serve non-human users; the first to support gigabit speeds to every connected device; yet also the first to simultaneously support low-bandwidth apps and devices consuming very little power; the first air interface built on network virtualization; the first to commercialize millimeter wave assets for consumer uses; the first to incorporate in a fundamental way the use of licensed and unlicensed spectrum.

But 5G also follows another pattern set in 4G, where the fundamental objective of the  mobile network is to support internet access, not voice. One has to search pretty hard to find any discussion of voice, in any discussion of 5G, simply because voice will be supported as just another IP or internet-based app. In other cases, voice will supplied by the 3G or 4G network.

Some might argue voice should be a bigger part of the 5G standard. Others will argue that voice can be handled by the 3G and 4G networks, and as voice becomes a feature, not a revenue driver, that low-investment approach arguably makes sense, as nobody expects voice to be a big or bigger revenue contributor in the 5G era.

Global harmonization of 5G spectrum, as with global harmonization of every mobile platform’s spectrum bands, or global air interfaces, is helpful to grow big markets and create network and consumer gear economies of scale.

On the other hand, there often are countervailing pressures, of the “market creates the standard” sort. That often happens when a large internal market, available resources and first mover advantages all align.

In other cases, the key enabler seems to be that significant spectrum is available for use, in regions adjacent to currently-used bands.

Right now, the best characterization is that 5G will use all kinds of spectrum, at many frequencies.

That is a pretty big change from past practice, where specific frequencies were allocated for specific network platforms. Much of the attention now centers on use of frequencies below 6 GHz, even if most of the additional spectrum will come in the ranges beyond 6 GHz.

GSA, for example,  believes 3300 MHz to  4200 MHz and 4400 MHz to 4990 MHz will be the primary spectrum between 1 GHz and 6 GHz for the introduction of 5G.

Parts of the band 3300 MHz to 4200 MHz and 4400 MHz to 4990 MHz are being tested:

• Europe                   3400 – 3800 MHz (awarding trial licenses)

• China                      3300 – 3600 MHz (trial), 4400 – 4500 MHz, 4800 – 4990 MHz
• Japan                      3600 – 4200 MHz and 4400-4900 MHz
• Korea                      3400 – 3700 MHz
• US                           3100 – 3550 MHz (and 3700 – 4200 MHz)

source: NTT

In the regions above 6 GHz, a number of bands are being looked at:

• USA:      27.5 GHz to 28.35 GHz and 37 GHz40 GHz deployments in 2018
• Korea:   26.5 – 29.5 GHz trials in 2018 and commercial deployments in 2019
• Japan:   27.5 – 28.28 GHz trials planned from 2017 and potentially commercial deployments in 2020
• China:    24.25 – 27.5 GHz and 37 – 43.5 GHz studies
• Sweden: 26.5 – 27.5 GHz awarding trial licenses for use in 2018 and onwards
• EU:        24.25 – 27.5 GHz for commercial deployments from 2020

In addition, the bands 600 MHz, 700 MHz, 800 MHz, 900 MHz, 1.5 GHz, 2.1 GHz, 2.3 GHz and 2.6 GHz may be of  interest for traditional and new applications in the internet of things (IoT) areas.

Why "Winner Take All" is So Common

Winner take all is the adage describing market leadership in most application categories, but also might be said to apply in capital-intensive industries such as the telecom business. Some might argue that structure describes most markets, over time, including retailing, transportation, and information technology.

The percentage of total revenue at publicly-traded U.S. corporations earned by the top 100 firms was 53 percent in 1995, for example, growing to 84 percent over the next two decades.

The difference share makes is clear in big markets, especially, according to researchers at Consultancy

Others would say the “winner takes all” market structure is particularly apt to occur in two-sided markets (transaction markets linking buyers and sellers; media markets or markets largely supported by advertising).

Some will note that markets with network effects also tend to produce the “winner take all” pattern, since such markets are built on value that increases disproportionately with the number of potential connections. In the old telecom business, that was exemplified by the “anyone can call anyone” level of scale.

It is easy to illustrate the fact that mobile app usage is highly concentrated on the supplier side, illustrating the winner take all pattern, where the leader in a category has 40 percent market share, or greater.

"The continuous growth of messaging and social apps mean that the total app time is becoming dominated by just a few sectors, with the top three categories accounting for 78 percent of all mobile app time spent," says Dr. Hannu Verkasalo, Verto Analytics CEO.

The point is that “winner take all” is surprisingly common, in many markets. In fact, it might be considered the “normal outcome” in any competitive market, according to analysis used by Constellation Research.

In many markets it is not uncommon for a few leaders to earn 70 percent of profits, if not revenue, for example. In fact, the phrase "winner takes all" normally refers to profit share, not revenue share or units shipped share.

Big Test for Ting in Denver Suburban Markets

If Google Fiber so far has proven to be the biggest single disappointment among the ranks of overbuilders and proponents of greater local loop competition, it is hardly alone. There are some other “for-profit” overbuilders and perhaps a couple hundred not-for-profit municipal networks.

Among the more interesting is Ting, owned by Tucows, a provider of internet domain name registration services.

And though would-be municipal overbuilds seem to be gathering momentum, most such networks have worked in rural areas and small towns.

Also, so far, it appears that most such efforts peak at between 20 percent and 25 percent market share. Overbuilder Ting says it intends to get 50-percent market share in five years in it markets, an aggressive target that likely never has been reached.

Ting says it has been able to get about 20 percent share initially, in a number of its internet access markets.

“We expect to see 20 percent adoption amongst serviceable addresses in a year and 50 percent in five years,” Ting CEO Elliot Noss always says. That latter forecast likely is quite important.

Noss talks about “network cost per customer” in somewhat oblique ways. There is cost per passing and cost per customer. The former figure does not change with adoption. The latter figure is highly dependent on take rates.

“At these take rates we will be paying about $2,500 to $3,000 per customer and those customers will be worth about $1,000 a year in margin,” says Noss (Noss probably meant “revenue,” not “margin.”). That makes sense at about 20 percent penetration and network costs per home passed of about $500.

Getting 50-percent market share really changes the “cost per customer” of the network. At 50 percent (every other home is a customer), the per-customer cost of the network drops to about $1,000.

“So we think about $1,000 to $1,400, plus the install to build a home,” he said. That seems to imply the cost per customer at 50-percent penetration, with about $1,000 in network costs and then about $400 for activating a drop and supplying customer premises equipment.

You're not going to get that $1,000 down to $500,” he says.

All that implies a network cost between $500 and $600 per passing. That would be in line with network construction costs to build the municipal network in Chattanooga, Tenn.

So the basic math for any overbuilder doing a fiber to home network involves network capex of about $500 to $600 per location, and about $400 per activated customer location.

Against that are a couple major assumptions, including take rates and average monthly recurring revenue. At 50-percent take rates, an overbuilder should have a terrific business, even with average revenue per account as low as $70 a month to $90 a month.

At 20 percent longevity could be a key issue. So, all other things being equal, the keys for overbuilder success are capex control, ARPU and take rate.

Significantly, Ting also is entering new suburban markets in Denver, Colo. that are bigger than what Ting has attempted in the past.

Ting Internet continues to gradually add locations in Charlottesville, a town of 46,000.

Ting also is offering commercial internet access service in Holly Springs, N.C. a community of about 30,000.

Early in 2017 Ting saw commercial activations in Westminster, Colo., a town of about 102,000 people. And Ting also will be building in Centennial, Colo., a suburb of about 120,000 people.

Local Access is Not a Natural Monopoly; But Might be a Somewhat-Natural Duopoly

It always has been difficult to challenge the incumbent U.S. telcos and cable companies in the consumer access services business. For that reason, hopes for sustained and widespread competition to the incumbents hinges on how feasible such challenges are, or might become.

We can say that in every market, local access is not a natural monopoly, as exemplified by the presence of three to four mobile operators, plus at least one fixed network retailer. In a few markets there is facilities-based competition between cable TV and telcos on a widespread or even ubiquitous basis.

What remains unclear is the natural progression of markets over time, such as how many mobile or fixed providers with their own facilities can sustain themselves long term.

“Overbuilding” of U.S. telcos and cable TV companies by third parties is not new, and has been going on for decades. But it has been a niche undertaking, for the same sorts of reasons business service providers have had to specialize.

Fixed networks are horrifically expensive, so few would-be competitors can raise capital to enter the market. And even if capital can be raised, large sums generally cannot be raised if the business model is a complete, head-to-head competition with both cable TV and telcos in a market.

The obvious result is that there are a limited number of cases where a facilities-based fixed networks attacker can create a sustainable business model.

Where such competition has proven sustainable, it is on a “niche” basis.

That has tended to be urban cores for alternate access providers (generally metro fiber companies), all-IP, high-capacity long-haul routes and selected large buildings for business services.

In the consumer space, high-rise buildings have been a logical target, followed by selective operations in suburban areas of big metro areas and small towns in rural areas. There is a simple reason for that pattern. Fixed networks are expensive and stranded assets are a big issue. So attackers tend to focus on niches.

“We are focused on efficient capital spending,”  says Wide Open West. In other words, the firm cannot build everywhere.

WideOpenWest, for example, “operates primarily in economically stable suburbs that are adjacent to large metropolitan areas as well as secondary and tertiary markets (smaller towns).”

The firm also grows incrementally by building in new areas adjacent to where it already has operations, using a technique known a edge-out, “making capital-efficient decisions and leveraging our existing operating infrastructure,” says WideOpenWest.

Even Google Fiber, with an ambitious overbuilder plan, found its neighborhood-by-neighborhood building program a challenge.

WideOpenWest, for example, ranks about sixth among U.S. consumer triple play providers, and might also be the largest overbuilder, with annual revenues of about $1.2 billion, passing about three million locations and claiming about 780,000 accounts, with take rates  about 26 percent, across all 300 communities in 19 markets.

source: WideOpenWest  

WideOpenWest overlaps Comcast about 53 percent of the time and competes with Charter Communications 39 percent of the time, and with AT&T virtually all the time. In some cases, WOW competes with Verizon FiOS (3.5 percent of cases) and Frontier Communications (2.7 percent).

Between the start of 2014 and end of 2016, though, the number of accounts served by WOW has declined slightly, as video and voice units have been lost, as internet access has grown.

That is not to say expansion (further edge outs) is precluded. Indeed, that has been a key growth driver for WOW. But WOW’s results might well suggest that a competent overbuilder has trouble sustaining market share beyond about the mid-20s level, over a long period of time, as markets evolve and incumbents create new sources of value.

Impact of Common Carrier Rules on Internet Access is Either Unknowable or Mixed

"What might have happened" with U.S. infrastructure spending, had common carrier rules not been imposed, is unknowable. "What did happen" is complicated.

It remains difficult to assess whether capital spending, under common carrier rules, grew, remained the same or shrunk, for several reasons. Some argue capex in the U.S. market grew, while others argue it declined. Different assumptions matter here.

Some point to actual levels of spending, and argue that is the proof of what happened. Others argue the issue is “what would have happened, if the rules had not been in place?”

As to what actually happened, spending was up and down, depending on industry segment and by company. So it is possible--perhaps likely--that multiple drivers were at work, making it impossible to sort out the common carrier impact from the other forces at work.

It also is impossible to prove what investment levels might have been, in the absence of the rules and contestant perceptions of what value the higher investment could have delivered.

Nor is it easy to describe the change in expectations created by anticipation of the rules. Although the common carrier rules were levied in 2015, industry observers had been expecting those rules for some time, causing investment hesitance, it is argued.

It might be reasonable to argue that expectations matter. It is logical to invest more heavily when a market with good profit margin is growing; less logical when markets are declining and have poorer prospects. It is logical to invest more when there is a guaranteed rate of return; less logical to invest heavily when markets are competitive.

It is logical to invest more heavily when there is a chance to significantly take market share; the opposite stance being logical when, despite investment, share might still be lost.

So it can be argued that there are clearly different dynamics at play in the U.S. mobile, fixed and cable TV segments, not a single trend.

At a high level, mobility capex tended to grow (growth was possible and competition increasing, so there was an investment push and a pull); fixed network capex was flat to declining (unclear whether revenue would increase, even if the investments were made); and cable TV capex was up slightly (cable could see a way to take a clear lead, and was gaining market share).

As a background trend, global telecom capital investment levels have been flat or declining since about 2000.

Further, investment sometimes is driven by reasons other than immediate expected return. In other words, investment might have been driven, in the current period, by considerations other than the common carrier rules.

I remember well being told, off the record, by a fixed network telecom executive, that the fiber ot home decision was, in fact, not driven by the usual investment considerations (invest A, earn a 20-percent return above return of capital) but by strategic concerns.

In other words, “we are investing to keep our business,” and not because the financial return really was expected to be positive in the normal sense. Given the heightened levels of telecom industry competition, some decisions were driven by competitive issues unrelated to common carrier rules, and, again, might well have been higher in the absence of the rules.

The point is that a number of drivers exist for capital investment decisions. It is hard to isolate the impact of common carrier regulations from the rest of the considerations. And it is impossible to determine what might have happened, had the rules not been in place.