AT&T Critics are Simply Wrong About Linear Video

Inevitably, aside from claims that the “wheels are coming off” the linear video business, there will be renewed criticism that AT&T should instead have spent the capital used to acquire DirecTV, and then (if approved by regulators) Time Warner, to upgrade its consumer access networks.

The critics are wrong; simply wrong, even if it sounds reasonable that AT&T could have launched a massive upgrade of its fixed networks, instead of buying DirecTV or Time Warner (assuming the acquisition is approved).

AT&T already has said it had linear video subscriber losses of about 90,000 net accounts in the third quarter. In its second quarter, net losses from U-verse and DirecTV amounted to about 351,000 accounts.

Keep in mind that, as the largest U.S. linear video provider, AT&T will lose the most customers, all other things being equal, when the market shrinks.

Some have speculated that AT&T potential losses could be as high as 390,000 linear accounts.

Such criticisms about AT&T video strategy might seem reasonable enough upon first glance.

Sure, if AT&T is losing internet access customers to cable operators because it only can offer slower digital subscriber line service, then investing more in internet access speeds will help AT&T stem some of those losses.

What such criticisms miss is that that advice essentially is an admonition to move further in the direction of becoming a “dumb pipe” access provider, and increasingly, a “one-service” provider in the fixed business.

That key implication might not be immediately obvious.

But with voice revenues also dropping, and without a role in linear or streaming subscription businesses, AT&T would increasingly be reliant on access revenues for its revenue.

Here is the fundamental problem: in the competitive era, it has become impossible for a scale provider (cable or telco) to build a sustainable business case on a single anchor service: not video entertainment, not voice, not internet access.

In fact, it no longer is possible to sustain profits without both consumer and business customers, something the cable industry is finding.

So the argument that AT&T “should have” invested in upgraded access networks--instead of moving up the stack with Time Warner and amassing more accounts in linear video with the DirecTV buy--is functionally a call to become a single-service dumb pipe provider.

That will not work, and the problem is simple math. In the fiercely-competitive U.S. fixed services market, any competent scale player is going to build a full network and strand between 40 percent and 60 percent of the assets. In other words, no revenue will be earned on up to 60 percent of the deployed access assets.

No single service (voice, video, internet access) is big enough to support a cabled fixed network. Period.

That is why all scale providers sell at least three consumer services. The strategy is to sell more units to fewer customers. Selling three services per account is one way to compensate for all the stranded assets.

Assume revenue per unit is $33. If one provider had 100-percent adoption, 100 homes produce $3,000 in gross revenue per month. At 50 percent penetration (half of all homes passed are customers), just $1650 in gross revenue is generated.

At 40-percent take rates, gross revenue from 100 passed locations is $1320.

But consider a scenario where--on average--each account buys 2.5 services. Then, at 50-percent take rates, monthly gross revenue is $4125 per month. At 40-percent adoption, monthly revenue is $3300. You get the point: selling more products (units) to a smaller number of customers still can produce more revenue than selling one product to all locations passed.

The point is that it is not clear at all that AT&T could have spent capital to shore up its business model any more directly than by buying DirecTV and its accounts and cash flow.

That the linear model is past its peak is undeniable. But linear assets are the foundation of the streaming business, and still throw off important cash flow that buys time to make a bigger pivot.

One might argue AT&T could have purchased other assets, though it is not clear any other assets would have boosted the bottom and top lines as much as did DirecTV.

What is relatively clear is that spending money to become a dumb pipe internet access provider will not work for AT&T, even if all the DirecTV capital had been invested in gigabit networks. At best, AT&T might have eventually slowed the erosion of its dumb pipe internet access business. It would not have grown its business (revenue, profits, cash flow) enough to justify the diversion of capital.

Would AT&T be better off today, had it not bought DirecTV, and invested that capital in gigabit internet access? It is hard to see how that math would play. Just a bit after two years since the deal, AT&T would not even have finished upgrading most of the older DSL lines, much less have added enough new internet access accounts to justify the investment.

AT&T passes perhaps 62 million housing units. In 2015, it was able to deliver video to perhaps 33 million of those locations. Upgrading just those 33 million locations would take many years. A general rule of thumb is that a complete rebuild of a metro network takes at least three years, assuming capital is available to do so.

Even if AT&T was to attempt a rebuild of those 33 million locations, and assuming it could build three million units every year, it would still take a decade to finish the nationwide upgrade.

In other words, a massive gigabit upgrade, nationwide, would not have generated enough revenue or cash flow to justify the effort, one might well argue.

Assume AT&T has 40 percent share of internet access accounts in its former DSL markets. Assume that by activating that network, it can half the erosion of its internet access accounts. AT&T in recent quarters has lost perhaps 9,000 accounts per quarter. Assuming AT&T saves 10 percent of those accounts, that amounts to only about 900 accounts, nationwide.

That is not enough revenue to justify the effort, whatever the results might be after a decade, when all 33 million locations might be upgraded.

source: Bruce Kushnick

The simple point is that AT&T really did not have a choice to launch a massive broadband upgrade program, instead of buying DirecTV, and instead of buying Time Warner. The financial returns simply would not have been there.

How Important is Network Effect in Telecom? Not Very, at Industry Level

It is a truism that network effects are key to many businesses and industries, ranging from social media networks to access networks. The network effect describes a situation where the value of a product or service is dependent on the number of others using it.

But there are other “scale” effects that shape the profitability of business models, ranging from “economies of scale” (lower unit costs possible because of customer mass or volume) to “economies of scope” (efficiencies created by product variety, not single-product volume).

source: atkearney.co.uk

Ideally, if one had a choice to operate in any industry, the “best” pick would be an industry with low scale requirements and high barriers to consumer switching behavior. Such industries are hard to find, as many consumer-facing  businesses require scale.

In that respect, internet access is in a difficult environment. It requires lots of capital, tends to earn lower profit margins than many other industries, requires some amount of scale and yet also has low barriers to customer switching behavior.

So scale does matter. Scale helps a supplier cope with high churn and high fixed costs.

There are other nuances. Economies of scale can be built on the producer side (firms get bigger) or the consumer side (more customers join a single network).

In the U.S. mobile service provider business, both producer and consumer network effects are important, but arguably less important than the scale advantages (scale and scope). In principle, every mobile customer, on any network, can communicate with all other customers of all other networks, meaning that the network effect for “mobility” is very high, even if, at the firm level, there is not such a network effect.

In other words, the universe of networks enjoys high network effects, even if no single small mobile service provider can gain such effects on its own.

Perhaps oddly, the network effect no longer confers mobile or fixed service providers much business advantage, since all suppliers must interconnect. On the other hand, economies of scale and scope do matter quite a lot.

Small telcos, cable operators and internet service providers now find they cannot make a profit on linear video subscriptions, for example. Likewise, service providers that offer only a single service (internet access, voice/messaging or video entertainment) will find their business models quite challenging.

New 2-Sided Markets?

Two-sided markets have existed in the content business for quite some time, even if new versions have appeared in the form of ride-sharing and room-sharing services. Such two-sided markets, with one salient exception, might be hard to create, in the mobile industry.

In the older version, a market maker makes money from both “buyers” and “sellers” of some product, even as the market maker primarily connects buyers (content consumers) with sellers (content owners).

Video subscription providers have earned revenue by selling consumers subscriptions (access) to desired content, but also have earned revenue from advertisers and content owners (local advertising the former; carriage fees in the latter case). More recently, content owners have turned the tables and now generally extract carriage fees (affiliate fees) from distributors.

But distributors still earn subscriber fees from consumers and local advertising revenues from third parties.

The strategic issue for at least some tier-one communications service providers is how to create new platforms, supporting new markets, that connect buyers and sellers.

Some of the new platform opportunities are more traditional “one-sided” markets. Consider mobile advertising, where revenue is earned only on one side of the platform, from third parties that want to place messages reaching the platform’s users (mobile customers).

That likely will be true for most internet of things use cases as well, where network operators will earn revenues from enterprises or other organizations that want to place sensors into the environment. Revenues in that case are earned by supplying the connectivity for the sensor networks, from “one side” of the platform.

At least in principle, two-sided markets are conceivable, where the mobile service provider also operates at the application level, owning and then operating marketplaces that connect buyers (job seekers, content seekers, patients) and sellers (employers, content suppliers, medical services providers).

source: digitalpolicy.org

Google Fiber Getting Out of Linear Video: No Surprise

The developing storyline around Google Fiber is that it is getting out of the linear video subscription  business because that business is dying. One need not look so hard for other conventional explanations. As every small telco and cable operator has known for decades, without scale, it is impossible to actual make a direct profit from linear video.

That was true even at the height of demand for linear video subscriptions. A couple of decades ago, as part of a due diligence review by a telco looking to buy cable TV assets, I asked the CEO what would happen to his business model if the subscription rate dropped from the then-current 90 percent rate to 70 percent. He immediately blurted out “then we’re dead.”

Keep in mind that this was before internet access or voice services were part of the bundle. The point is that even for a cable company with hundreds of thousands of subscribers, there was very little room in the business model for a dip in subscribers from 90 percent to a lower figure.

For any firm such as Google Fiber, with less than 100,000 subscribers, the video entertainment business case (as a stand-alone product) almost certainly will not work. For video as for any other consumer service, scale really does matter, a fact small cable TV and telco operators always have lived with.

Peak Mobile? Data is Mounting

It is too early to tell whether one sign of a peak product life cycle has been reached in the U.S. mobile market. It is possible a current bout of price competition will ameliorate and reverse as the U.S. market consolidates.

Also, mobile operator data data revenue growth was negative, for the first time ever, in the first quarter of 2017. That matters because such growth has driven overall revenue growth for 17 consecutive years.

Altogether, about $3 billion worth of quarterly revenue has disappeared from the U.S. mobile service provider market in the second quarter of 2017, representing a $12 billion annual loss, if the trend continues.

Verizon suffered its first ever decline in service revenues, year over year, according to analyst Chetan Sharma. 

Also, for the first time, net adds for connected (cellular) tablets were negative as well. For the first time, while postpaid net-adds were negative as well. 

To be sure, T-Mobile US has taken market share and revenue. But the bigger story is the revenue shrinkage.

source: Business Insider

To be sure, T-Mobile US has taken market share and revenue. But the bigger story is the revenue shrinkage.

source: Business Insider

How Much Market Share Can Independent ISPs Take from Incumbents?

How much market share can independent internet service providers take from cable companies, telcos and mobile operators? In the market as a whole, not so much. The business remains a matter of scale, and scale is expensive.

On a local level, the impact can be quite significant, in principle. In the mobile market, about two percent share is held by all firms other than the four national leaders.

Ting Internet has a rather dramatic set of assumptions for deciding where it enters a new internet access market: it assumes it will get 20 percent market share at launch, growing to 50 percent market share within five years, in markets where a telco and a cable operator already operate.

Aggressive? Yes. Even Verizon, where it has built and marketed fiber to the home services for years, typically gets market share only in the 40-percent range.

To be sure, the U.S. internet access market represents perhaps $160 billion in annual revenue, of which fixed access generates nearly $60 billion annually. But scale is hard to achieve in market so big.

Also, there is some evidence that users are switching to mobile internet access. Fixed network internet access subscriptions in the United States have declined in recent years, falling from 70 percent in 2013 to 67 percent in 2015, for example.

Some 13 percent of U.S. residents rely only on smartphones for home internet access, one study suggests. Logically, that is more common among single-person households, or households of younger, unrelated persons, than families. But it is a significant trend.

Some suggest that service providers are actively pushing mobile services as an alternative to fixed access, for example.

In fact, some studies suggest that U.S. fixed internet access peaked in 2009, and is slowly declining, though other studies suggest growth continues. Still, some studies suggest U.S.  fixed network subscriptions declined in 2016, for example.

source: Pew Internet

In the second quarter of 2017 alone, some 228,000 net new fixed network access accounts were added. Ting probably has about 3,650 total internet access accounts, and generates just over $4 million worth of annual revenue.

source: pwc

So Ting assumes it can, within five years or so, essentially relegate one or more of the two dominant providers to a barely-profitable or unprofitable status. The issue is how big a force Ting eventually could be, since capital is a huge constraint to its achieving meaningful scale in the fixed network internet access business.

Unlike some municipal or government network models, Ting does not assume it will get financial support from one or more universal service funds or a shift of existing government communications spending.

The Ting payback model also includes a fiber access network construction and then customer attachment cost of $2,500 to $3,000, with each customer expected to contribute $1,000 or so per year in gross revenue.

As others now do when building gigabit internet access networks, Ting also builds first in neighborhoods where it believes demand is highest, as demonstrated by customer deposits.

Some question the sustainability of the business model. So far, internet access arguably drives revenue growth for the company. It is fair to note that Tucows remains a small company, booking annual revenues of perhaps US$336 million.

Almost by definition, Tucows does not have the financial ability to take much total market share in the U.S. internet access market, dominated by firms with scores of billions to hundreds of billions in annual revenue.

Vertical Integration is Risky, But Might be Imperative

Vertical integration now is becoming an important strategic issue in the applications and communications industries, welcome trend or not.

Whether or not most mobile and fixed communications operators are able to move up the stack,  Google is clearly moving down the stack into devices, retail internet access, undersea capacity and new access platforms. It is not alone.

Amazon’s purchase of Whole Foods, the grocery store, moves Amazon elsewhere in the distribution chain.

Amazon’s creation of its own air freight operation, and now its moves into retail package delivery provide other examples of app layer integrating backwards into the value chain.

source: wikiwand.com

And Amazon long ago got into the devices business (KIndles, Echos, phones, tablets).

In the video entertainment business, content networks and owners are moving to integrate content distribution platforms to go direct to consumers as well.

Apple designs its own integrated circuits.

As always, the point of such vertical integration is to capture business benefit. "Vertical integration is simply a means of coordinating the different stages of an industry chain when bilateral trading is not beneficial,” say McKinsey consultants.

Generally speaking, value chains that feature adjacencies where there are few sellers and few buyers create business risk.

Reliance on a single, or a few customers, is a source of business risk because of the implications of losing those few customers, or having them significantly reduce buying volume. Likewise, reliance on a single, or just a few products, creates similar risk.

For Google, internalizing undersea capacity assets simply saves it money, while providing better quality control and flexibility. Amazon finds the same motivations drive its creation of air freight and local delivery services.

For Google, creating its own devices supplies the same value as does Apple in tightly integrating software and hardware. On the other hand, intervening in the internet access markets causes all the traditional suppliers to upgrade the existing access infrastructure.

Few firms arguably succeed at vertical integration. But successful moves can create barriers to market entry by other firms. And that is why vertical integration matters.

AT&T believes it can, and likely must, vertically integrate. At the moment, its moves to become a content producer and owner are the most-obvious example. In the future, moves to integrate various internet of things applications, services and platforms are likely to be equally important.

Simply put, the mobile services business model has reached saturation in the U.S. market, and firms such as AT&T must find entirely new services to sell to its customers.

You would be correct in arguing that such vertical integration is highly risky, and that horizontal acquisitions make more sense for most firms, when possible. The problem is that horizontal growth sometimes is difficult to impossible.

In its home market, horizontal expansion is not possible for AT&T, for regulatory reasons. Horizontal expansion is possible internationally, but will not help decline in its core market, absent some significant vertical movement.  

For other firms, such horizontal acquisitions take lots of capital, and that generally means additional debt. That can be difficult in a business that normally requires fairly high levels of debt, in any case.