Business Context Shapes Access Network Strategy

As often happens in any industry, service providers have different opinions about fixed wireless access versus fiber-to-premises versus hybrid fiber coax versus satellite platforms for access services. 

As always, different firms have different views on strategy because of their business circumstances. Perhaps in principle, all former telcos would say fiber-to-premises is the ideal long-term solution where the economics exist. But the economics are daunting in many cases, leading to a “yes, but” strategy that uses other platforms as the economics dictate. 

Verizon has a relatively small “in region” footprint of U.S. homes and businesses--perhaps no more than about 20 percent--and cannot afford to “fiberize” another 80 percent of U.S. homes. So fixed wireless, which piggybacks on the 5G network, makes sense. 

T-Mobile, with close to zero fixed network coverage of U.S. homes and businesses, benefits even more from 5G fixed wireless. 

AT&T, on the other hand, has the biggest footprint of U.S. homes and businesses, so out-of-region coverage magnitudes are correspondingly reduced. 

Comcast and Charter have “homes passed” totals close to AT&T’s footprint and already have HFC networks they believe will be marketplace competitive for quite some time, as multi-gigabit speeds are coming next on the HFC platform. 

Of a total of 140 million U.S.  homes, AT&T’s landline network passes 62 million. Comcast has (can actually sell service to) about 57 million homes passed.

The Charter Communications network passes about 50 million homes, the number of potential customer locations it can sell to.

Verizon homes passed might number 27 million. Lumen Technologies never reports its homes passed figures, but likely has 20-million or so consumer locations. 

Assuming no further significant consolidation, AT&T only “needs” to fiberize within its footprint to reach 44 percent of U.S. homes (and virtually all the homes regulators are likely to allow it to pass). 

Assuming Verizon has no appetite to significantly expand its fixed network footprint, that leaves about 81 percent of U.S. homes that could be passed by the 5G network, and would be impossible to significantly serve using FTTH. 

Comcast already has HFC offering gigabit speeds reaching about 41 percent of U.S. homes. Charter already passes about 29 percent of U.S. homes. Again, regulators are unlikely to allow either firm to get significantly bigger, in terms of homes passed. 

Lumen has a largely-rural territory that includes perhaps 14 percent of U.S. homes, but Lumen has no mobile network assets it can use to offer fixed wireless on a facilities basis. 

The point is that each firm’s view of strategy is shaped by its existing legacy assets. Cable operators, though not denying FTTH makes sense in the future for a growing percentage of customers, also believe HFC is a viable platform, without major reliance on FWA or FTTH to serve mass market customers. 

Verizon and T-Mobile have good reasons for using FWA that piggybacks on their nationwide 5G networks. 

AT&T believes FTTH is the best solution, but also has the largest in-region fixed network footprint of any major ISP, and therefore has the most to lose if copper access facilities are not upgraded to fiber access.

There is no universal answer for “which access platform” makes most sense. Each major ISP has key business model constraints and opportunities that shape the access network choices.

Headline Home Broadband Speeds and Real-World Buying Patterns

It always is instructive to compare advertised headline speeds with the services customers actually buy, as the mismatch is often striking. Despite all the talk of “gigabit per second” and “multi-gigabit per second services, most customers do not buy them. 

This analysis by Ofcom of the actual speed plans U.K. customers were buying between 2018 and 2022 shows that most customers were purchasing speeds in the range of 30 Mbps to 100 Mbps for the whole period. 

                      Home Broadband Advertised Speeds Take Rates

    Nov '18 Nov '19 Nov '20 Mar '21 Mar '22

10 Mbps or less         2%     1%     1%     0%     0%

<10 Mbps >30 Mbps 33%     24%     15%     15%     9%

<30 >100 Mbps         48%     56%     60%     60%     65%

<100 Mbps >300         16%     16%     20%     19%     19%

300 Mbps or more         1%     3%     4%     5%     8%

source: Ofcom 

As you might expect, there has been a shift towards higher speeds over time. One finds the same general pattern in the U.S. market: about 15 percent of customers buy services operating at 1 Gbps or faster; about seven percent take services offering 500 Mbps to 900 Mbps; while 55 percent buy service at speeds between 200 Mbps and 400 Mbps. 

source: OpenVault  

Global average speeds are lower.

source: Hootsuite 

Many estimate that by 2025, the “average” home broadband user might still require less than 300 Mbps worth of capacity. Nielsen’s law of course predicts that the top available commercial speeds in 2025 will be about 10 Gbps. 

source: NCTA 

But “headline” speed services will not be the products most home broadband consumers buy.

FTTH is Important, But Only So Important

One might be forgiven for overestimating the value of fiber-to-home networks; fixed network revenues in general or profit margins from fixed network business services. After all, the fixed network once drove 100 percent of revenue and all profit. 

Nobody seems to contest the notion that fiber-to-premises networks are the future. And an awful lot of investment activity now goes into building and acquiring access networks that are largely copper-based but could be transformed into fiber access facilities. 

It is undoubtedly the case that optical fiber, deployed quite deep into access networks, is the intermediate requirement for fixed network survival and relevance. Just as arguably, all that has to be kept in perspective.

For all the investment, and for all the value, fixed networks just do not generate all that much revenue in the connectivity business. It might seem that is the case to practitioners, but it is largely an illusion, in the context of the global business. 

The global connectivity business now is driven by mobile operations. 

Consider AT&T. Mobility generated most of the revenue and even more of the actual profits. At AT&T, about 11 percent of total revenues are generated by all consumer fixed network operations. 

Revenue generated from business customers on the fixed network represent about 19 percent of total revenues. So the fixed network contributes about 30 percent of total revenue. 

Mobility accounts for 70 percent of total revenue. 

source: AT&T data, Daniel Jones formatting 

source: AT&T data, Daniel Jones formatting 

Just as significantly, of the roughly $25.8 billion in profit, only about five percent of total profit was generated by all fixed network operations. 

Fully 95 percent of profits are generated by the mobile operations.

source: AT&T data, Daniel Jones formatting 

At Verizon and T-Mobile, mobility plays an even-bigger role. T-Mobile has virtually nil fixed network revenue, while Verizon in 2019 stopped reporting fixed network revenue altogether.

Perhaps that will help put FTTH into perspective. In a real sense, FTTH now is a niche within the global connectivity industry, because fixed network access operations are a niche within the industry. 

That is not to deny that, in any given segment, that segment is close to 100-percent of the importance of the connectivity business for those who are in that part of the business. 

Still, as important as the FTTH investment thesis remains, it might affect less than five percent of consumer service profits. The fixed network represents about 15.5 percent of total profits, with business customers contributing about 11 percent of the profit total. 

The point is that, despite all the important decisions service providers and investors have to make around FTTH, the whole fixed network supplies a smallish portion of total revenue. The fixed network remains important, but pales in comparison to the mobility business as a driver of revenue and profit.

Sometimes an Industry Has to Raise Prices

Home broadband customers do not like price increases anymore than do mobile subscribers or video subscription customers. But, by definition, high inflation rates mean a general rise in virtually all prices. 

On the other hand, U.S. high-speed home broadband prices arguably have declined over the past couple of decades, especially in relation to prices of other products. 

According to BroadbandNow, U.S. home prices “have fallen since 2016, with the highest speed plans falling the most.”

The average price decreased by $8.80 or 14% for 25 – 99 Mbps. The average price decreased by $32.35 or 33% for 100 – 199 Mbps. The average price decreased by $34.39 or 35% for 200 – 499 Mbps. The average price decreased by $59.22 or 42% for 500+ Mbps, according to BroadbandNow analysis. Other studies show the same trend. 

According to the Blandin Foundation, prices dropped between 16 percent and 38 percent between 2015 and 2020, for example. 

source: Blandin Foundation 

Data from the U.S. Bureau of Labor Statistics shows prices for health insurance, for example, rose 43 percent between 2015 and 2021 while home broadband prices dropped 26 percent. 

sources: BLS, High Speed Options formatting 

The same downward trend was seen in Australia home broadband markets. 

sources: BLS data, NCTA formatting

ISPs sometimes argue their business models are unsustainable if capital investment keeps rising and average revenue per account  keeps dropping. 

There is a “simple” if difficult solution: raise prices to cover costs. As difficult as that is in competitive markets, there is no other long-term and sustainable solution. Subsidies can help, but only so much. Revenue has to be greater than cost. 

Consumers will not like it, but retail prices have to rise.

Price's Law and Organizational Incompetence

Price's Law states that half of the literature on a subject will be contributed by the square root of the total number of authors publishing in that area. In principle, it is similar to the Pareto theorem, which states that 80 percent of outcomes are produced by 20 percent of the actions. 

Extrapolated to organizational output, Price’s Law suggests 10 percent of people produce half the outcomes, while 90 percent produce the other half. 

source: Darius Foroux 

And remember it is a square root or power law function: the disparities grow larger with scale. The percentage of people producing half the value actually decreases with scale. The Pareto theorem, for example, is linear. It suggests 20 percent of actions produce 80 percent of value, at any scale. 

source: ANG Traders 

Price’s Law is different. As population size grows, though the number of those contributing half the value grows, they grow at ever-decreasing rates in relation to the total number of associates. 

source: Semantic Scholar 

That is one reason why very-large organizations contain so many people who are apparently not functioning at a high level.

Lower Prices are a Feature, Not a Bug, for Policymakers

Lower prices are a policy feature, not a bug. Government policy promoting competiton is designed to create lower retail prices. So it should not be surprising that pressure on average prices per user or customer or account now are a major service provider concern.

What else would you have expected? Lower prices are the intended outcome of competition policy.

To what extent is it correct to characterize legacy service provider revenue trends as “down and to the right?” Obviously legacy services such as fixed network voice, mobile messaging and voice and linear entertainment video generally show that pattern. 

What we often forget is that the very objective of introducing competition for connectivity services, and the government policies to support that objective, are intentionally designed to create lower prices. The objective of policy is a “down and to the right” pattern. 

In other words, “down and to the right” is not a bug, it is a feature. It represents the outcome policy intends, and is not a defect of policy. 

source: United Nations

The other angle is that a proven way of increasing ARPU is to increase speed, despite another clear trend: over time, speeds grow but prices remain relatively flat, or even decline. In other words, the cost of a 300-Mbps connection is the same, or less, than a 512-kbps connection three decades ago. 

Over the short period of 2007 to 2017, for example, U.S. typical speeds grew by two orders of magnitude, while prices dropped. 

At the moment, up to 80 percent of U.S. locations can buy internet access operating at least the gigabit per second level.  

source: Versa Technology 

The global pattern is not always transparent. If one looks only at total service provider revenue, that tends to grow each year, in large part because new customers are added in growing regions (Asia, primarily, but also eventually in Africa). 

So global service provider revenues will grow 14 percent between 2022 and 2027, according to researchers at  Omdia. Monthly average revenue per user will fall by four percent. So the pattern is more customers, each paying less than the “typical customer” used to pay. Revenue might ber “up and to the right” but ARPU clearly is “down and to the right.”

source: Omdia 

Every management team seems to emphasize that value can be enhanced, preventing further commoditization. You can make your own assessment of how effective such efforts have been, or could be. 

Disposable income is among the other limitations. It will be hard to boost ARPU very much in lower-income countries. If policymakers succeed in reducing the cost of connectivity to perhaps two percent of gross national income per capita, that further limits ARPU upside. 

source: S&P Global Market Intelligence 

In other words, the goal of policy in developing countries is to actively reduce ARPU. The objective is precisely “down and to the right” pricing. 

Until recently, home broadband was the major product line producing “up and the right” results, but growth now has slowed in mature markets. That pattern looks more like “flat and to the right.”

When service provider executives talk about a transition from “telco to techco” they essentially are saying such moves will change the revenue picture to “up and to the right.” 

Service providers in emerging or younger markets have advantages, in that regard. They can still hope to rely on mobile subscription growth and uptake of mobile internet access to fuel their continued “up and to the right” growth prospects.

Mature market executives who own infrastructure assets have no such luxury, and strategic options often hinge on whether mobility revenues exist. Contestants in mobile or fixed businesses who operate using wholesale access, rather than owning infra, have other options. 

Competitors with Infra-based business models complain about higher capital investment requirements and limited abilities to monetize those investments, not without basis in fact. 

Still, we often forget that the whole point of introducing competition in access markets is to drive average costs “down and to the right.” It is a policy feature, not a bug.

Are FTTH Payback Models Sustainable? What's Good for Private Equity Might Not be So Good for Operators

Some of us would admit to being surprised at the payback models  for fiber-to-home deployments, the degree of business moat protection some believe FTTH offers, and therefore the value of such digital infrastructure assets, compared to other assets such as cell tower sites, data center or edge computing assets. 

Looking back over 25 years of business model assumptions, it is quite startling how much the underlying assumptions have changed. Subsidies now play a bigger role, offering in some cases a 20-percent to 30-percent reduction in capital investment in rural markets. 

On the supply side, though demand is not altered, private equity investment now means more capital is available to accelerate build timetables. 

But the most-shocking change are the revenue assumptions for consumer locations. These days, the expected revenue contribution from a home broadband account hovers around $50 per month to $70 per month. Some providers might add linear video, voice or text messaging components to a lesser degree. 

But that is a huge change from revenue expectations in the 1990 to 2015 period, when $150 per customer was the possible revenue target. In some cases, revenue up to $200 per home location was considered feasible. 

Expectations now hinge almost exclusively on consumer home broadband. 

“Our fiber ARPU was $61.65, up 5.3 percent year over year, with gross addition intake ARPU in the $65 to $70 range,” said John Stankey, AT&T CEO, of second quarter 2022 results. “We expect overall fiber ARPU to continue to improve as more customers roll off promotional pricing and on to simplified pricing constructs.”

Lumen reports its fiber-to-home average revenue per user at about $58 per month.

Recent presentations also have shown fiber-to-home home broadband average revenue per user of about $63. 

source: Frontier Communications 

Granted, most larger ISPs believe they can boost ARPU over time, by adding features, adding speed tiers and moving customers to higher-priced plans with higher usage allowances. 

Market share or installed base is the other huge assumption. Can most providers expect to get 20 percent take rates or as much as 50 percent? And what other assumptions about operating cost are necessary to create a sustainable business case at 20-percent share? 

Most incumbent telcos deploying FTTH have been able to get 40-percent market share after several years of marketing. But is a terminal rate around 40 percent to possibly 45 percent (or even 50 percent) reasonable in most cases? If not, where is that possible? 

On the supply side, capital investment benefits from government subsidies and to some extent infra cost declines, though construction costs are stubborn. So while some larger ISPs hint at per-passing network costs as low as $600, others report costs closer to $1,000 per passing, with connection costs of $550 to $600 per customer. 

source: Matt Nicholson Lewis 

The point is that even with subsidies, lower infra gear costs and new investment sources, the demand expectations for consumer services have been slashed as much as two thirds over the past several decades. Many ISPs no longer expect revenue contributions from voice or entertainment video sources, and must build their demand models based solely on internet access. 

The largest ISPs might still expect some revenue contribution from voice or video services, but seem to be modeling higher expectations for business connectivity or contributions to mobile infrastructure cost models. In other words, the cost of small cell infra is aided by the consumer FTTH investment. 

Ultimately, we will see whether  FTTH really underpins a business that provides a competitive moat, while throwing off predictable cash flow. It seems more likely that private equity could succeed in transforming a legacy copper-based access business into a fiber access business, with a boost in equity multiples that will justify the effort.

That might well be a different question than asking whether FTTH really is a real estate or infrastructure asset on par with airports, toll roads, electrical and gas utilities. Much of the bet relies on limited competition. The argument that the first FTTH provider in a suburban or urban market gets 40 percent market share is likely correct. 

That has a reasonable chance of being correct even if two equally competent providers operate their own FTTH networks in a market. 

Some believe the first-mover advantage in a rural market could be substantial enough to support higher market shares. 

But it remains a valuable exercise to ask whether the FTTH business model is sustainable on an operational business on the revenue scales now being seen. 

That is a different question than asking whether a private equity buyer can boost multiples--and then sell the asset--by replacing copper access with optical fiber access. 

As a matter of operating economics, it still seems unclear whether FTTH networks generating only $50 to $70 per month in residential revenue are sustainable, assuming legacy provider cost structures. A small, lean upstart should have an easier time, as embedded costs are lower than those found at legacy firms.