Facebook, Google Are, Will Be, ISPs. Will Apple or Amazon Follow?

Few observers would be shocked if Facebook moved ahead with plans to create a satellite-based Internet access service aimed at billions of potential users in the global south. Facebook has talked about using satellites and drones as ways to deliver affordable Internet access services.

Facebook also has discussed other methods, including low earth orbit satellites, and has been working on free satellite-delivered Internet service in Africa.

Boeing Co. said it could get a deal in 2015 to build a high-throughput communications satellite for at least one leading application or device supplier.

The type of satellite Boeing talked about would be deployed in geostationary orbit.

Who the buyer might be is the big question. Some suggest it could be Facebook, Google, Apple or Amazon, or perhaps discussions have occurred with all four firms.

Google already has purchased satellite manufacturing assets, and invested $1 billion in SpaceX , which has announced plans to launch a new fleet of satellites to provide Internet access. Some might note that the Google investment appears to be in SpaceX as an entity, not just the new fleet of satellites.

And Google is testing a variety of methods for supplying Internet access across the global south, including drones, balloons and apparently, satellites.

Separately, OneWeb plans a huge fleet of new low earth orbit satellites to provide Internet access services.

Jim Simpson, vice president of business development and chief strategist for Boeing Network and Space Systems, says big technology firms have direct financial interests in expanded Internet access.

"The real key to being able to do these type of things is ultra high-throughput capabilities, where we’re looking at providing gigabytes, terabytes, petabytes of capability," Simpson said.

Both Google and Facebook have talked about use of satellites as part of efforts to bring affordable Internet access to perhaps billions of people who cannot afford to buy it, or cannot buy it because affordable service provided by terrestrial networks do not reach them,  right now.

It would come as no surprise if Facebook emerged as the buyer, since Facebook has been most open about use of geosynchronous satellites for Internet access.

Some would deem Google a  “not surprising” but also not “most likely” buyer. The big shocker would be if a firm such as Amazon or Apple were the buyers.

Should the deal happen, within a couple of years at least one big app provider would emerge as a competitor to existing satellite retailers and mobile service providers.

It isn’t clear whether most observers would see Amazon or Apple as potentially more dangerous competitors than Google or Facebook, if only because observers expect to see both Google and Facebook get into the Internet access business, in some form. Indeed, Google already is an operating ISP in the U.S. market.

On the other hand, it is conceivable that either Facebook or Google could wind up being partners, to an extent, with mobile operators, though potentially competitive as well. Much depends on whether the new satellite-based ventures are retail or wholesale oriented, or at least what the balance is, between retail and wholesale operations.

That is almost an expectation. What would be more unexpected is an entry by Amazon or Apple into the satellite Internet access business, in some form.

When are Consumer Markets Effectively Competitive?

When is a market effectively competitive? And what are the key tests of whether markets are effectively competitive? To examine the matter simply, is it the number of suppliers in a given local market, or the market shares of those competitors?

Of course there is the argument that it is a combination of those two metrics. But the answers might hinge on which test is dominant.

The U.S. Federal Communications Commission, for example,  seems to apply different standards to different services within the triple play bundle.

In high speed access, the Commission appears to believe there is not enough competition, with two terrestrial providers and two satellite providers in just about every market.

In the case of linear video services, the FCC thinks competition is effectively competitive with the same number of suppliers in virtually every market.

So perhaps it is the market share test that is relevant.

Still, the market share held by the market leader in most local markets does not vary much, whether looking at high speed access or linear video.

Broadly, a cable operator has about 59 percent share of the high speed access market, while in linear video the local cable TV operator tends to have about 52 percent share.

What differs is the share held by other contestants. In high speed access markets, though there are a growing number of competitors (two satellite Internet providers, Google Fiber an independent ISPs), market share generally is held by the cable TV operator and a local telco.

On a national basis, a cable provider might have 59 percent share, the telco might have 40 percent share. Satellite and other providers have the rest of the share. Local markets will vary much more widely, however.

Still, high speed access tends to feature four providers in most markets, despite the fact that the market share structure is functionally a duopoly.

In linear video, market share is much more dispersed. Nationally, cable operators have about 52 percent share, while satellite providers have about 36 percent share, while telcos have about 12 percent share.

In the linear video market, virtually every market has four providers, as well. In 1993, a cable operator had about 93 percent share.

The linear video market arguably is more competitive, one might say. On the other hand, should AT&T succeed in buying DirecTV, the number of suppliers in linear video, in most markets, would drop to three.

The result would be that in many markets, where AT&T has fixed network operations, though hat a cable operator might still have roughly 52 percent share, AT&T might have 27 percent share.

Dish Network would still have perhaps 15 percent share. In some markets, where AT&T does not have local fixed network facilities, Verizon might have about 25 percent to 30 percent share of linear video.

More competition is coming, in voice, high speed access and linear video markets. But the FCC now seems to have concluded that voice and linear video now have become effectively competitive, though it believes high speed access and mobility have not reached that stage, yet.

What Shape Will Internet Access Price Regulation Now Take?

The Federal Communications Commission continues to insist it will not apply price regulation to the Internet interconnection and access businesses as part of its decision to regulate Internet access and transport under common carrier rules.

With the caveat that enforcing a “zero price” rule is price regulation, should the network neutrality rules survive legal challenge or legislative override, unexpected price obligations might still be incurred by edge providers, ironically enough.

Though we are far from knowing precisely how the rules will be interpreted, if the common carrier framework survives legal challenge, there are all sorts of ways prices now might be regulated.

Zero pricing is the reality created by the "best effort only" or "no fast lanes" portion of the rules. That is the price at which consumer ISPs can price use of their networks by edge providers (apps). 

On the other hand, under common carrier rules, interconnecting networks pay compensation for termination whenever traffic loads are not equal. By definition, content domains (app and content providers) impose far greater, and highly unequal, termination traffic on "eyeball" domains (ISP customer bases).

It is one thing for the FCC to claim it will "forebear" from imposing price regulations. 

It might be quite another matter if the common carrier rules on network interconnection or termination apply as they have in the past. In that case, the FCC might be unable to prevent normal network termination charges from being applied, when traffic flows clearly are unequal.

Why Hybrid Cloud Now is Service Provider Focus

Despite the growth of public cloud operations, private cloud will represent 69 percent of cloud workloads in 2018, says Kelly Ahuja, Cisco SVP. That essentially explains service provider interest in hybrid cloud services.

"When people discuss cloud, they often focus on public cloud services or public cloud storage services,” Ahuja said. “Even with public cloud workloads having significant growth, by 2018, almost 70 percent of cloud workloads will still be private cloud-related.”

According to Gartner analysts, use of hybrid cloud will triple over the next three years. The percentage of organizations that prefer a hybrid approach will grow from seven percent to 20 percent in 2017, for example.

Put simply, there is a growing business for suppliers who are able to support hybrid cloud operations that bridge private and public cloud domains. No matter what the specific capacity or access provider involvement in cloud computing (simple transport, hosting, access, integration or cloud services of the software, infrastructure, platform, business process variety), hybrid operations are the largest potential segment, one might argue.

By 2018, 69 percent (113.5 million) of the cloud workloads will be in private cloud data centers, down from 78 percent (44.2 million) in 2013, and 31 percent (52 million) of the cloud workloads will be in public cloud data centers, up from 22 percent (12.7 million) in 2013, Cisco says.

Cloud operations also are growing as a percentage of total data center traffic. In 2013, cloud accounted for 54 percent of total data center traffic, and, by 2018, cloud will account for 76 percent of total data center traffic.

Through 2017, the average enterprise network will see a 28 percent compound annual growth rate(CAGR) for bandwidth due to the use of cloud computing, mobile devices and video, according to IDG.

By 2017, enterprises that do not control network use risk requiring up to 3 Mbps per user of

committed bandwidth, or more than 20 times the average need in 2012.

The bottom line is that capacity requirements, not to mention related integration and professional services, plus apps generated by hybrid cloud arguably arguably represent the single most accessible segment of cloud services revenue for most transport and data center providers.

Slowly, Google is Becoming a Major ISP

Up to this point, Google has been a direct retail Internet service provider only in the U.S. market, though there have been rumors Google is thinking about doing something similar in Europe.

Google now has announced it is launching a new mobile virtual network operator service in the United States, and most observers expect the intent is to test better ways of integrating Wi-Fi and mobile access. To the extent those efforts center on Long Term Evolution and Wi-Fi, the implications are global.

Years ago, Google began to build its own internal global undersea network. Opinions varied about whether there were wider implications. As it turns out, Google has indeed become a commercial ISP in the United States. Whether it will do so elsewhere is the new issue.

And though Google Fiber is a fixed network business, Google now is experimenting in mobile as well.

And it is certain Google will deepen its role in the Internet access business globally. Google recently made a $1 billion investment in a commercial satellite Internet access venture. And that is not all.  

Google’s fully-owned drone manufacturer--Titan Aerospace--has gotten two test licenses from the U.S. Federal Communications Commission, allowing Titan to test “Internet by drone” services for five months.

The tests will happen inside a 520 square mile area to the east of Albuquerque, from March 2015 to September 2015. The testing area would include Moriarty, where Titan Aerospace is based.

Nobody outside Google knows precisely what Google might ultimately do, and where, with a range of wireless and mobile Internet access platforms.

Google recently made a $1 billion investment in a venture to deliver satellite Internet access services.

In addition to ownership of Titan Aerospace, Google is testing the use of unmanned balloons to deliver Internet access.

At least three mobile service providers are testing the platform, which interworks with a Long Term Evolution mobile network. Were those efforts to become commercial operations, Google would operate as a backhaul provider for mobile service providers, as well as a retail service network.

Google also owns Skybox Imaging, a manufacturer of satellites.

Then there is the mobile virtual network operator business Google is launching in the United States, plus Google Fiber, the gigabit fixed network Internet access business, plus Google’s operation of Wi-Fi networks for U.S. Starbucks locations, and its investment in mobile apps of many types.

In the past, Google has made investments in Clearwire, the former U.S. mobile services company, municipal broadband, Android operating system and Nexus phones and tablets.

Nobody knows which of these efforts will emerge as significant commercial operations and which are intended primarily to influence wider market adoption of practices Google considers helpful to its own business.

Should Google commercialize even a few of the new initiatives, it would emerge in a variety of potential roles, including backhaul provider, retail Internet supplier, or both. In such cases, Google might be a partner for some ISPs, a competitor for backhaul suppliers or a competitor to retail ISPs.  

And, it goes without saying, Google is both emerging as a potential new competitor to other access providers, a partner and supplier. In all cases, though, Google's business objective is to reduce barriers to the use of the Internet and the cost of Internet access. 

That means, among other things, that Google has a vested interest in lower cost smartphones, computers and tablets, as well as lower cost Internet access, both mobile and fixed. Always.

Uh Oh. The "Bubble" Word Reemerges

Those who lived through the telecom bubble and bust know viscerally what it means when trillions worth of telecom assets suddenly evaporate. So many will be chilled to hear suggestions that the U.S. market might be in a position similar to the pre-burst conditions of the middle 1990s.

Between 1997 and 2003, the global telecom industry lost $2.8 trillion of market value, collapsing from $4 trillion to $1.2 trillion, an event virtually without parallel in the industry, and arguably in any industry.

Some might argue that irrational behavior occurred because capital markets relied too heavily on Federal Communications Commission rules and regulations, either favorable or unfavorable.

Massive capital misallocation--and huge losses--in a very short time, was the result.

Ironically, some might argue, we now are about to see the reverse of the irrational over-investment, namely rational underinvestment, but with a similar causation: federal telecommunications regulation.

The argument is that with profit potential in the U.S. high speed access market, leading contestants will find they can earn a better return elsewhere than in the U.S. market.

Common carrier regulation in general, and network neutrality rules in particular, will limit supplier business models and prospects for revenue growth in the U.S. high speed access market--at least for many of the tier one suppliers.

Supporters argue the gap will be filled by new suppliers, operating with lower costs. That essentially is what happened when cable TV operators, for example, successfully attacked the U.S. fixed network voice and high speed access markets.

That seems already to be happening, on a small scale, as independent Internet service providers launch new gigabit service operations on a local basis. Eventually, given enough scale, such new competition will pressure tier one provider business models.

The point is that major shifts in U.S. telecom policy within the last two decades have lead to drastic changes in investment, competition and innovation, with what can only be called huge distortions of capital allocation.

Some might note the irony that the passage of telecom legislation intended to increase competition lead to greater dominance by the incumbents. In one sense, that might be correct. In the wake of the Telecommunications Act of 1996, hundreds of new competitors arose, only to crash and burn by 2003.

In other other vital sense, competition did increase, however, as U.S. cable TV companies became major players in voice and high speed access, arguably now taking the lead in the strategic high speed access category.

What now happens will be worth watching. Investment is going to change. The only issue is how much, and where that investment will shift. AT&T’s moves in Mexico might provide a bit of insight.

More significant will be Verizon’s response, since Verizon has bet its near term future on robust U.S. market growth. If that should change, so will Verizon’s strategy.

Cloud Computing Drives Global Bandwidth

“Cloud computing” now is becoming so fundamental that it is driving international bandwidth patterns and growth. Since data centers are central to cloud computing, it likewise will come as no surprise that data centers have functionally displaced central offices as the network nodes that drive traffic across networks.

There also is a good reason why “hybrid cloud” is viewed as a big opportunity by many in the cloud ecosystem. Despite the growth of public cloud operations, private cloud will represent 69 percent of cloud workloads in 2018, says Kelly Ahuja, Cisco SVP.

"When people discuss cloud, they often focus on public cloud services or public cloud storage services,” Ahuja said. “Even with public cloud workloads having significant growth, by 2018, almost 70 percent of cloud workloads will still be private cloud-related.”

By 2018, 69 percent (113.5 million) of the cloud workloads will be in private cloud data centers, down from 78 percent (44.2 million) in 2013, and 31 percent (52 million) of the cloud workloads will be in public cloud data centers, up from 22 percent (12.7 million) in 2013, Cisco says.

Cloud operations also are growing as a percentage of total data center traffic. In 2013, cloud accounted for 54 percent of total data center traffic, and, by 2018, cloud will account for 76 percent of total data center traffic.

Over the next five years, Cisco forecasts a tripling of data center traffic. Consumer Internet usage will drive much of the increase.

By 2018, half of the world's population will have residential Internet access, and more than half of those users' (53 percent) content will be supported by personal cloud storage services.

Mobile usage likewise will drive traffic. The Asia Pacific and North America regions will account for a little over half of global mobile traffic by 2019.

But the Middle East and Africa regions will experience the highest CAGR of 72 percent, increasing 15-fold between 2013 and 2018.

Central and Eastern Europe will have the second highest CAGR of 71 percent, increasing 14‑fold over the forecast period.

Latin America and Asia Pacific will have CAGRs of 59 percent and 58 percent, respectively.

Between 2013 and 2019, mobile traffic will grow 11 times in the Middle East, Central Europe and Africa, according to Ericsson.

Through 2018, the Middle East and Africa region is expected to have the highest cloud traffic growth rate (54 percent compound annual growth rate [CAGR]); followed by Central and Eastern Europe (39 percent CAGR); and Asia Pacific (37 percent CAGR).

Between 2009 and 2013, Middle East International bandwidth demand grew 61 percent compounded annually, according to TeleGeography.

One might well make the argument that revenue shares within the broader information technology business could shift, as a result of the new computing architecture.

Canalys reports that the global IT industry represented more than US$2 trillion in spending in 2012, for example. All “cloud” spending was seven percent of that amount, or about $14 billion.

Public cloud services delivered by Amazon Web Services, Rackspace, Microsoft, Google and others accounted for just four percent of IT spending.