Wednesday, January 26, 2022

AT&T FTTH Metrics Improve, But Deployment Pace and Take Rates Really Matter

In its competitive battle for home broadband customers, AT&T’s fortunes depend on three key drivers: fiber to home coverage; take rates and any average revenue per account gains that could supply, with the primary variable being coverage. 


The reason is simply that AT&T cannot challenge dominant cable TV providers for installed base and market share until the company has much-greater FTTH coverage. Simply put, AT&T and most other local exchange carriers cover too few homes to go head to head with cable home broadband. 


To be sure, AT&T and other telcos are pushing FTTH deployments at an accelerated pace. AT&T expects to have 30 million home locations passed by FTTH by about 2025, up from about 15 million to 15.5 million at the moment. 


source: AT&T 


Keep in mind that AT&T passes a total of about 57 million homes. So the company’s current FTTH coverage is between 26 percent and 27 percent of its total passings. 


AT&T will not be able to go head to head with cable, across the full range of home broadband speeds, until it has FTTH available to most homes in its fixed network footprint. By 2025, when AT&T expects to  have 30 million FTTH passings, it will be able to compete head to head in about 45 percent of its footprint. 


Take rates are arguably the second most-important variable, as there is a difference between an active account and a location able to buy. Over the past year, AT&T has boosted its FTTH take rate from 34 percent of passings up to 37 percent of passings. 


The goal is to approach 50 percent take rates, which would exceed the general take rate of about 40 percent telcos have been able to garner over the past couple of decades. 


Finally, it remains to be seen how average revenue per account changes as more customers take FTTH home broadband services. At the moment, AT&T’s bottom tier FTTH home broadband service (exclusive of taxes) runs about $55 per month. The mid-range tier costs about $65 a month, while the top gigabit tier sells for $80 a month.


In addition, AT&T has added two premium tiers offering 2-Gbps symmetrical and 5-Gbps symmetrical access for $110 and $225 per month, respectively. The mix of accounts could strongly affect AT&T revenues. 


The lowest FTTH tier--offering 300 Mbps--covers more than half of the U.S. home broadband buyer base. Another 17 percent of customers buy services operating between 200 Mbps and 400 Mbps. 


source: Openvault  


Assuming lower-income households take advantage of support programs, those households reached by AT&T could have internet access at 300 Mbps for about $25 a month. 


That is why the pace of FTTH upgrades matters so much for AT&T. To reach parity with cable TV operators, which AT&T defines as market share greater than 50 percent and installed base approaching 50 percent. Without nearly-ubiquitous FTTH deployment, those goals are likely unreachable. 


To be sure, competition at the lower speed levels--up to about 100 Mbps, is an opportunity for Verizon and T-Mobile. Verizon has a limited fixed network footprint of U.S. homes while T-Mobile has had zero share of the home broadband market. 


Even if those two firms use fixed wireless to reach the lower and middle tiers of buyers, that is about 65 percent to 70 percent of today’s home broadband market. 


For AT&T and other telcos, though, the pace of FTTH deployment will be the story.


U.S. FTTH Payback Model is Changing

This chart showing AT&T fixed network average revenue per user for AT&T consumer connections might surprise you. 


At the end of 2020, the chart shows, AT&T had about 14 million broadband customer accounts in service and had some five million FTTH connections. 


That might be low. AT&T itself reported it had 5.5 million FTTH accounts in service at the end of the third quarter of 2021.  


Other estimates for the same period suggest AT&T FTTH passed 15.5 million home locations in late October of 2020. “Passings” mean a customer can buy. AT&T says it had a take rate for FTTH passings of 34 percent in 2020. 


In 2021 take rates seem to have climbed, and might have reached 37 percent.  

source: Digtl Infra 


Keep in mind that AT&T had about 57 million total homes passed in its fixed network footprint in 2021. Even if AT&T spends less than $1375 per active account location, and perhaps in the $500 to $600 range for passing (but not activating) a home, the payback model almost certainly has to rely on a number of subtleties.


In some cases, AT&T may already have usable fiber-to-node facilities in place, which should reduce the cost of extending fiber to the home. It is possible that in some locations government subsidies could amount to perhaps $300 per line. 


But some of the financial value also has to be attributed to enterprise, small business and small cell deployments that can use much of the same distribution fiber. Presumably some of the payback will come from lower operating expense as well. 


Still, with overall ARPU for combined copper, FTTN and FTTH facilities at just $58 a month, once has to assume some amount of voice and video entertainment revenue is involved. 


At a high level, AT&T has been saying FTTH payback models work at $50 a month ARPU and penetration of 50 percent. 


If one assumes that the mainstay has to be home broadband, with average household broadband spending of about $40 to $50 a month, with digital subscriber line unsustainable longer term at about $40 a month, then one is almost forced to assume that a substantial portion of AT&T’s FTTH revenues (stand alone, with zero voice or video contribution) will be closer to $60 a month. 


AT&T FTTH charges for 300 Mbps service already are at $55 a month (ignoring taxes). Service at 500 Mbps costs $65 a month and 1-Gbps service retails for $80 a month.  


Even assuming every FTTH customer only buys service at 300 Mbps (more than half of all U.S. homes buy service between 100 Mbps and 200 Mbps), revenue with zero contribution from voice or video would be above $50 a month.


Tuesday, January 25, 2022

Will Connectivity and Cloud Resources Eventually be Sold Through Exchanges?


Many ideas in the connectivity business do not catch on immediately. Some of you may remember Enron Broadband and its effort to create a bandwidth trading exchange. That attempt failed, partly because of criminal behavior, partly because the value proposition for potential partners was not clear, partly because all the costs of creating liquid exchanges had not developed to the point where it was easy and affordable. 

But Moore's Law operates. All of the computing resources we need get cheaper with time. Open source has become ever more present as a networking concept. 

Cloud computing and connectivity have grown increasingly intertwined. Networks are becoming virtualized. Newer tools such as blockchain change the certainty and ease of settlements between trading partners. 

Where we might well be headed is the creation of connectivity and cloud resource ordering and provisioning that is very much akin to the way e-commerce sites work generally. 

It should not come as a surprise. The goal of networking for 50 years has been "bandwidth on demand." Now computing is available on demand. 

Perhaps nobody in the business denies, in principle, that over time, we will move towards services and applications on demand. 

This is a precursor. 

Physics Operates--Though it Seems Impossible--When "Noseriding"

Some things we know are possible, such as “noseriding a long surfboard, have physics that seemingly remain unexplained. Since a rider’s center of gravity on a longboard is normally somewhat back of the center of the board, a nose ride visually looks as though it is the prelude to a disaster. 


Obviously, physics is at work, allowing one hundred to 150 pounds--or more--to be on the extreme leading edge of a nine-foot board that otherwise requires a balance point center to slightly center rear. 


Still, it is possible, and happens all the time. It is the Coanda effect, achievable only when the board is moving relatively fast, on a wave that is moving relatively fast. You cannot do this on a slow, mushy wave. 


In other words, much as an airplane wing creates lift, the Coanda effect creates enough lift to allow a rider to do something that otherwise would result in falling off as the board dips its nose into the water and ends the ride.


The design of the board can enhance the ability to nose ride, creating more lift under the nose and drag on the tail that counterbalance the apparently-unbalanced position of the nose rider. 


source: Reddit 


5G Private Networks Do Not Necessarily Displace Wi-Fi

Private 5G networks--where an enteprise uses 5G as a local area network platform--are a "new use case" for 5G networks, beyond the obvious support of the latest-generation mobile network. But different industries seem to have different appetites for private 5G.


Nor, in principle, does a private 5G network necessarily displace the need for Wi-Fi. As virtually all observers note, private 5G might have application where process controls, privacy concerns or extremely low-latency applications must be supported.


In many instances, private 5G will also be used because local computing is desirable, typically for reasons of latency performance, high compute intensity or bandwidth needs as well. So demand for private 5G networks tends to correlate with edge computing, in many cases.


In other cases, network slicing using the public network might also be favored.


In the new Real Wireless white paper How Ready is the UK for 5G? It is clear that private network use cases predominate across a range of industries, especially manufacturing and healthcare. 


Transport and logistics as well as creative industries show a mix of private 5G and public 5G modes. 


In manufacturing, private networks are expected to dominate. 

Source: Real Wireless


Healthcare shows a different pattern: private networks delivered using a public network.


Source: Real Wireless


Transport and logistics shows a mix of private and public modes. 


Source: Real Wireless


Creative industries likewise show a mix of private and public networks. 


Source: Real Wireless


Monday, January 24, 2022

AT&T Launches Commercial 2-Gbps Symmetrical and 5-Gbps Symmetrical Fixed Network Internet Access

As predicted by Edholm’s Law, internet access speeds continue to climb. AT&T, for example, just activated symmetrical 2-Gbps and symmetrical 5-Gbps service for 5.2  million locations across 70 U.S. markets, with plans to deploy across the whole footprint in 2022 and later years. 


Edholm’s Law states that internet access bandwidth at the top end increases at about the same rate as Moore’s Law suggests computing power will increase. 


Nielsen's Law, like Edholm’s Law, suggests a headline speed of 10 Gbps will be commercially available by about 2025, so the commercial offering of 2-Gbps and 5-Gbps is right on the path to 10 Gbps. 

source: NCTA  


Headline speeds in the 100-Gbps range should be commercial sometime around 2030. 


How fast will the headline speed be in most countries by 2050? Terabits per second is the logical conclusion. Though the average or typical consumer does not buy the “fastest possible” tier of service, the steady growth of headline tier speed since the time of dial-up access is quite linear. 


And the growth trend--50 percent per year speed increases--known as Nielsen’s Law--has operated since the days of dial-up internet access. Even if the “typical” consumer buys speeds an order of magnitude less than the headline speed, that still suggests the typical consumer--at a time when the fastest-possible speed is 100 Gbps to 1,000 Gbps--still will be buying service operating at speeds not less than 1 Gbps to 10 Gbps. 


Though typical internet access speeds in Europe and other regions at the moment are not yet routinely in the 300-Mbps range, gigabit per second speeds eventually will be the norm, globally, as crazy as that might seem, by perhaps 2050. 


The reason is simply that the historical growth of retail internet bandwidth suggests that will happen. Over any decade period, internet speeds have grown 57 times. Since 2050 is three decades off, headline speeds of tens to hundreds of terabits per second are easy to predict. 

source: Nielsen Norman Group, Future Speaker

 

Of course, the typical consumer customer does not typically buy the top-available speed at any given point of time. 


AT&T Fiber 2 GIG costs $110 per month plus taxes. AT&T Business Fiber 2 GIG costs $225 per month plus taxes. The “typical” consumer household probably now pays about $50 a month for internet access. 


AT&T Fiber 5 GIG: $180 per month plus taxes. AT&T Business Fiber 5 GIG costs $395 per month plus taxes. 


The point is not that so many consumer households will buy the top offers. The point is that--with higher “fastest” speed tiers, the typical buyer also tends to move up. As more customers buy the 2-Gbps or 5-Gbps services, a change in the adoption rate will happen when we hit about 10-percent adoption of either service. 


Adoption of 1-Gbps service now has passed the 10-percent point, so we should see accelerated adoption of gigabit services. Where half of U.S. consumers now buy services in the 100 Mbps to 200 Mbps range, we should see a shift of those buyers to higher-speed services over the next few years, as the top end also continues to move. 


Once 2-Gbps adoption hits about 10 percent of households, half of the rest of the market will start to move to speeds more in the range of 400 Mbps. Once 5-Gbps adoption hits about 10 percent, half the market will start a move towards gigabit service, history suggests.


LIke it or Not, "More Freedom" Might be the Only Way to Deal with Perceived Unfairness

The U.S. Supreme Court, in 376 U.S. 254 (1964) NEW YORK TIMES CO v. SULLIVAN established a principle that public officials cannot claim damages for “libel” against media entities for defamatory falsehoods “relating to his official conduct unless he proves that the statement was made with "actual malice."


In other words, a plaintiff cannot sue because a media entity published something found to be incorrect, but only if the published material was created with “malice.” It is not enough that something we might consider “libelous” was said; only that it was said intentionally, with malice and a “reckless disregard” for the facts of the matter. 


When a statement concerns a public figure, according to the Court, it is not enough to show that it is false for the press to be liable for libel. Instead, the target of the statement must show that it was made with knowledge of or reckless disregard for its falsity.


Hence the difficulty public officials (or public figures generally) face when suing a media outlet for libel: the plaintiff must prove malicious intent, not simply that something is “false.” 


Of course, such Constitutionally-based protections and rulings were developed before social media.  Whether social media calls for a rethink is an issue. Should social media be covered by the same rules as “media?” Is the “malicious intent” requirement too broad? 


More broadly, does the First Amendment protection of free speech only apply to government suppression of free speech, or to suppression by other means? Right now, social media are considered covered by the First Amendment in terms of all content moderation. They are classified as “speakers” for purposes of First Amendment protections. 


Perhaps there is no other way to consider them. Even if they act in ways many consider biased or unfair, there is a First Amendment argument that this is their right. 


And earlier government efforts to mandate a “Fairness Doctrine” have generally been considered failures, and even unnecessary in an era where so many outlets for speech exist. Mandating “fairness” might prove more onerous or ineffective than not mandating it. 


Still, media law and First Amendment law have evolved with the emergence of new media. The initial responses have often been to “limit freedom.” Over time, law has generally moved in the direction of “more freedom.” 


Even if not considered by many to be “fair,” this might be the only Constitutionally-correct approach: more freedom.


Will AI Disrupt Non-Tangible Products and Industries as Much as the Internet Did?

Most digital and non-tangible product markets were disrupted by the internet, and might be further disrupted by artificial intelligence as w...