Invest or Harvest? It Often Depends on Network Platform

In an industry that changes so rapidly, it often is useful to remember just how much has changed in telecommunications, globally, in a few short decades, and how busincess choices are affected by the scale and pace of change.

The broad context is that service providers have to make key decisions about when to harvest, and when to invest, in various products in their portfolios, as end user demand changes.

Consider use of fixed line voice, across a range of nations.

In the United States, customers buying fixed network voice now number about 42 percent of the population. In Italy, that figure is just 37 percent.

In the United Kingdom, customers number 59 percent of the population, 60 percent in France and 45 percent in Germany. Some might argue the numbers are that high in substantial part because of rules on line bundling related to purchases of high speed access, namely that purchase of a voice line is required to buy high speed access.

The main observation is that fixed network voice is a product in the declining phase of its adoption cycle. That poses key questions for suppliers.

How much more innovation should be attempted for fixed network voice? How much should harvesting” be the dominant strategy?

Some will argue for increasing investment to change the value proposition and therefore allow price increases. Others might argue for harvesting the legacy revenue as long as possible.

Contrast that with mobile adoption (measured in terms of active accounts), which stands at 106 percent in the United States, 159 percent in Italy, 140 percent in Germany, 117 percent in France and 130 percent in the United Kingdom.

Those adoption figures might suggest a similar “harvesting” strategy, but that would not be correct. Mobile, unlike fixed networks, are replaced about every decade. For that reason, a first generation or second generation network sells a different product than a third generation or fourth generation network.

In the mobile services sphere, “investing” in the next generation–not harvesting–is the typical, and arguably correct strategy.

In developing countries, strategics are just as clear.  In India, fixed voice adoption is just two percent of the population, but mobile adoption is at 91 percent. In Brazil, where fixed voice lines are purchased by 22 percent of people, mobile is bought by 137 percent of people.

The point is that it typically does not make sense to invest very much in fixed networks, when a similar investment in mobile will produce outsized revenue gains.

Mobile voice is approaching a peak of adoption in many countries, even in many developing nations. So how does the “invest or harvest” decision play out?

In some ways, Voice over LTE can be viewed as an investment in “next generation” voice.

In other ways, VoLTE can be viewed as a cost-saving measure, allowing operators to eventually decommission 3G networks (often used to support voice services) nd redeploy 3G bandwidth to support 4G.

Mobile Internet, on the other hand, remains a major growth opportunity in developing nations, while some forms of mobile Internet access, in most developed markets, already are nearing a peak.

Mobile Internet access adoption is 100 percent in the United States (again, measured in terms of active subscriptions), 113 percent in Japan, 130 percent in Sweden and 110 percent in South Korea.

Mobile Internet access (using third generation platforms, generally)  has a bit more growth left in other countries, including the United Kingdom, where adoption is at 77 percent, France, at 64 percent adoption, Germany, at 51 percent adoption and Italy at 74 percent take rates.

But it is fourth generation Long Term Evolution that will be the potential growth driver in most developed countries. LTE represents four percent of U.K. and French mobile Internet accounts, five percent of German accounts, three percent of Italian and Spanish accounts and four percent of Netherlands mobile accounts.

In the United States, LTE represents 23 percent of all mobile accounts in service, in Japan 22 percent, in Austria 20 percent. In India and China virtually all the growth lies ahead.

Fixed high speed access connections represent a harder-to-describe situation, as service is purchased “per location,” not “per person.” The key implication is that adoption “per person” is not as relevant measure, as the issue is more accurately “adoption per household.”

Fixed broadband adoption “per person” was 36 percent in the United Kingdom, 38 percent in France, 34 percent in Germany, 23 percent in Italy and 29 percent in the United States. But consumers buy fixed network high speed access “by location,” not “per person.”

In 2012, for example, U.S. household adoption of high speed access was 73 percent. As a percentage of population, high speed access stood at 29 percent. Obviously, “per capita” adoption doesn’t accurately reflect the number of people able to use the services.

In 2013, high speed access per household stood at 93 percent of households in South Korea, 90 percent in Switzerland, 83 percent in Japan, 82 percent in Canada and 75 percent in the United States.

In other words, “per capita” or “per person” measures of fixed network high speed access do not accurately portray the true state of adoption.

The point is how much has changed over just a couple of decades. Fixed network voice, the driver of industry revenue, already is becoming a relatively limited driver of total revenue, as the product is purchased by less than half of all households.

Meanwhile, revenue gains are paced by Internet access, video entertainment and mobile services. In some cases, it still makes sense to measure adoption and revenue on a “per account or per household” basis. Fixed network Internet access and video entertainment provide examples.

Increasingly, though, the more-important metric is revenue per user, revenue per account or revenue per device, as more services are sold in the mobile domain.

But revenue per account also makes more sense for fixed network services when the primary retail packaging is a bundle of three or four services.

And traditional “invest or harvest” decisions have a different context in the mobility business, where whole networks are replaced about every decade. For a mobile service provider, it will always make sense to invest in the next generation. Harvesting is a temporary expedient until the older platform can be retired.

source: Akamai

source: Ofcom

Why Quadruple Play?

In one way, it is obvious why telecom and cable TV service providers in many markets are shifting product lines and retail offerings to quadruple play bundles. Doing so allows the service providers to boost gross revenue and average revenue per account, while reducing churn.

Quadruple play offers also tend to provide a better platform for growing market share and therefore boosting subscriber numbers.

With such clear advantages, the grand strategy of offering a bundle of fixed network and mobile services requires little explanation.

Still, even when contestants agree on the value of the bundling strategy, larger firm growth imperatives can be quite disparate. Verizon Communications, for example, has bet $130 billion on its ability to grow almost entirely within the U.S. market.

AT&T, with its bid to buy Iusacell, the Mexico-based mobile service provider, at a minimal cost of about $2.5 billion, is following through on its different growth strategy, which is expansion into international markets.

That strategy has been in place for some time. AT&T reportedly considered a purchase of Telefonica, but Spanish regulators quickly moved to signal disapproval.

America Movil and partner AT&T also were rebuffed in 2007 in an effort to buy Telecom Italia. More recently, there were rumors that AT&T was considering buying Vodafone, a huge transaction that would have created the globe’s biggest communications service provider, at least in terms of market capitalization.

The $48 billion AT&T bid to buy DirecTV, on the other hand, is an important, but more limited gambit to create a nationwide quadruple play offer, something virtually no other service provider can do, at present.

To be sure, any U.S. service provider attempting to offer quadruple play services nationally would have to do so using wireless networks (satellite, mobile, untethered), as no single firm can afford to build a nationwide fixed network, even if the Federal Communications Commission would allow it–and the FCC will not allow it.

If and when Verizon Communications puts a national strategy into place, it also is likely to lean heavily on the mobile network, though likely with an emphasis on over-the-top streaming video rather than linear video. That is one area where AT&T and Verizon disagree on near-term strategy.

AT&T believes linear video still has legs. Verizon is less convinced, at least where it comes to its own involvement in entertainment video. Specifically, Verizon argues the skimpy profit margin from linear video will eventually be made worse when OTT streaming becomes a widespread alternative.

Still, there is another reason why many leading service providers are turning to quadruple play bundles. Competition and product substitution in the fixed network voice, mobile voice, mobile messaging, linear video (in some markets) and even high speed access (in some markets) product lines is making the whole business more challenging.

Whatever synergies might exist between fixed and mobile network assets, the simple fact is that revenue growth in some markets now virtually requires the revenue lift a quadruple play can provide, simply because, in a growing number of markets,  some or all of the revenue sources are stagnant or declining.

In other words, “scale” (selling more units) is becoming difficult to impossible, within each product segment. “Scope” (selling different products) therefore emerges as the rational strategy, within each market.

So far, AT&T and Verizon also disagree on how much room remains for market share gains in the original market. AT&T, which has grown primarily through acquisition, sees future growth hinging on additional acquisitions outside the traditional core market.

Verizon, which generally has grown organically–even though Verizon itself was formed from the merger of Nynex and Bell Atlantic–does not yet see the same need to expand internationally.

If Verizon strategic thinking changes, the reasonable prediction will be that international expansion will come primarily on the mobile side of the business. Both domestically and internationally, telecom industry revenue growth is driven by mobile services.

Nearly 80 percent of AT&T’s pre-tax income during the first six months of 2014 came from its mobile business. At Verizon, the mobile segment contributed 68 percent of revenue  in the third quarter of 2014 and 50 percent of earnings (earnings before interest, taxes, depreciation, and amortization).

Globally, mobility has been the revenue growth driver for decades. In 1998, mobile accounts were about a fourth of all voice lines in service globally. By 2001 mobile had just about reached parity with fixed, and by 2002, and ever since, mobile lines have surpassed fixed voice lines in service.

By 2009, mobile voice accounts in service stood at 67 percent adoption, compared to less than 18 percent adoption of fixed voice service, globally.

In fact, use of fixed voice lines peaked in developed markets in 2001, and had begun declining by 2002. In the U.S. market, use of fixed voice lines peaked in 2000.

Globally, use of fixed lines peaked in 2006, even if use in developing markets grows slowly.

But the mobile market is maturing as well. According to researchers at Ovum, global mobile connections will grow by a compound annual growth rate of less than four percent between 2012 and 2018, while global revenues will grow at less than half that rate.

Average revenue per user also will decline. Global mobile connections will grow from 6.5 billion in 2012 to reach 8.1 billion by 2018, while annual mobile service revenues will rise from US$968 billion to US$1.1 trillion.

However, global service revenues will contract in 2018 for the first time in the history of the mobile industry, declining from 2017 levels by one percent or US$7.8 billion, Ovum predicts.

Does that mean service providers will abandon expansion by acquiring out of region mobile assets? Not likely. It does mean they will pick their opportunities, though.

The ability to offer quadruple play services will be a more-important part of the screening process, in many cases.

by Gary Kim

In Europe or U.S., Investment and Competition Sometimes are Inversely Related

Predictably, AT&T has been criticized in some quarters for pausing gigabit access network investment until certainty about high speed access regulation is settled. To be sure, all participants in regulated, or potentially regulated markets, take actions to bolster their chances of winning an argument. That is as true for Google as it is for AT&T.

Uncertainty or regulations that reduce return on invested capital have a direct impact on deployment of next generation networks, many would argue. And it appears we need look no further than Europe for proof.

Incumbent market share across the EU-27 in 2010 was 38 percent. In other words, competitors had gained 62 percent share of the fixed network market.

But that degree of competition has come at a price. The EU is in danger of failing to make its announced goal of 30 Mbps by 2020, a target that originally was set before the launch of Google Fiber, which has changed market dynamics and investment in the U.S. market, for example.

So regulators should ease up on IT and telecommunications companies to allow them to compete with rivals around the world, said Guenther Oettinger, new European Union digital economy commissioner.

“So far, we have ensured that consumers benefit from the liberalization of telecoms markets,” Oettinger said.  From now on our actions must be more geared more toward allowing companies to make fair profits.”

That represents a huge change in thinking. The main point is not that the EU has decided to take a “North American” or “U.S.” approach. Instead, the big shift is the recognition that promoting competition and promoting investment can become rivalrous and mutually-exclusive goals.

In theory, regulators try to create regulatory frameworks that simultaneously promote both competition and expedited investment in next generation networks. In practice, almost any set of policies will be criticized.

That was the case in the United States prior to the Unbundled Network Element Triennial Review. Competitive local exchange carriers were thrilled about their ability to get wholesale access to incumbent telco switching and access facilities at healthy discounts.

The incumbent telcos predictably were unhappy.  As AT&T Chairman and CEO Michael Armstrong said in a 1998 speech, “no company will invest billions of dollars to become a facilities-based broadband service provider if competitors who have not invested a penny of capital nor taken an ounce of risk can come along and get a free ride on the investments and risks of others.”

After a rules change that eliminated mandatory access at highly discounted rates, competition in the consumer segment of the U.S. voice and high speed access market shifted to a war between U.S. cable TV operators and telcos, both of which owned and operated their networks.

Some competitive providers made incremental gains where they were able to focus on business customers, not consumer accounts.

Some will point to the Verizon FiOS upgrade as a direct response to the change in wholesale rules. Verizon Communications did not begin the massive investment in fiber-to-home facilities until after the 2003 rules change that ensured Verizon it would not have to sell wholesale access to FiOS except at negotiated commercial rates.

Others might note that the unregulated cable TV industry, by 2003, had gained a two-to-one advantage over telcos in high speed access market share, a development some observers attribute to the differences in regulation of telco and cable TV wholesale rules, between 1999 and 2003.

Simply, cable TV operators have never  been required to sell wholesale access to their networks. Between 1996 and 2033, telcos were required to do so at significant price discounts to retail. After 2003, when the mandatory wholesale rules were lifted, high speed access market share quickly equalized in two years.

To be sure, non-facilities-based competitors were not happy about the 2003 triennial review changes, as the switch in policy favored the cable TV companies, who had facilities in place, and essentially eliminated the profit for consumer service providers using wholesale access.

But some would say the switch to an emphasis on “facilities-based competition” has succeeded, as per-capita investment in U.S. access networks has been substantially higher than in the EC region; about twice as high, by some estimates.

It now appears that a decisive change in thinking has happened. EC authorities are worried about lagging investment in next generation networks, and are prepared to take steps to promote investment, not just competition.

The larger point is that policies matter. Both competition and investment in next generation networks are important. EC regulators now are thinking about how to promote investment, not just promote competition.

That remains a valid framework in the U.S. market as well, as the network neutrality framework is debated. Uncertainty jeopardizes investment. So do rules that jeopardize financial return from long-lived capital investments.

by Gary Kim

Why Telcos and Cable Companies are Relatively Sanguine About Dwindling Voice Revenues

The fact that few in the U.S. telecommunications or cable TV business are panicked (concerned, working on the problem, but not terrified) over continual declines in end user buying of fixed network  voice services or linear video is testament to the existing and future replacement products service providers have been able to create.

In the past, voice revenues have driven all, or most, communications service provider revenue. In many emerging markets, that remains the case. But voice has ceased to do so in the U.S. market.

Likewise, entertainment video has driven cable TV revenues. But all that is changing.

Many fixed network providers (cable TV and telco) already earn more money from Internet access than from voice services.

Already, AT&T earns $33.6 billion annually from its “data” category, which includes all U-verse video revenues. Voice services generate about $20.3 billion annually, for example.

In most markets, mobile voice revenues still surpass data. But few might contest the notion that, globally, mobile data revenue will exceed voice revenue by 2018.

Many might be surprised to learn that already is the case in a number of markets.

In 2012, Japan became the first country where mobile data revenues exceeded voice revenues. That also happened in 2013 in Argentina.

In the United States, mobile data revenue surpassed voice revenue in the first quarter of 2014.

That likely will happen in the United Kingdom in 2014 as well, according to the GSMA.

What might be more surprising, though, is that video might even surpass voice revenue, at least for some firms. AT&T provides a case in point. But European telcos including Vodafone and Telefonica are themselves growing the percentage of video revenue in their portfolios.

Assume video accounts generate $960 annually ($80 a month per video account), and that AT&T has 5.5 million such accounts. That implies annual revenue of about $5.3 billion in video revenue.

But DirecTV alone earned $31.8 billion in 2013.

Should AT&T succeed in its bid for DirecTV, video entertainment would possibly reach $37.1 billion, eclipsing even data services–at about $28.3 billion in annual revenue–as drivers of AT&T fixed network segment revenues.

After a DirecTV acquisition, voice would be only the trailing third most important revenue source for AT&T’s fixed network segment. Of $89.7 billion in total revenues, voice would represent 22 percent of fixed segment revenues.

Video would represent 41 percent of total fixed network revenues. Internet access and other data services would represent 32 percent of total fixed network revenues.

The picture at Verizon might be different. Verizon had $30.8 billion in first quarter 2014 revenue. About $18 billion was generated by mobile services, and though

About half of Verizon’s fixed network segment revenue comes from enterprise or wholesale sources.

So $17.3 billion is earned from “mass markets,” including small business. Removing small business accounts, the consumer business generates about $14.7 billion annually.

Verizon has about 5.3 million FiOS video customers. Assuming the same $80 a month contribution from a video account, Verizon might earn $960 per account, or about $5.1 billion from video services.

So video would represent about 35 percent of consumer segment revenue.

Assuming only $480 in annual revenues from Internet access, Verizon might generate about $4.3 billion from consumer Internet access, or 29 percent of consumer fixed network revenues. Voice might contribute about 36 percent of total fixed network revenues earned from consumers.

So, at Verizon, video entertainment and voice might be about equal contributors of revenue, while Internet access provides less revenue, in the consumer segment.

But keep in mind that the whole consumer segment is only half of fixed network revenue. Overall, voice and data revenues are much more significant at Verizon, than at AT&T.

It is a sign of how much has changed in the telecom business that voice might soon be the third largest revenue contributor, not the biggest.

A new study by Gallup illustrates why that is the case.

“One of the most striking cultural and social changes in the U.S. in recent decades has been the revolution in the ways Americans communicate,” says Gallup.

Texting, using a mobile phone and email messages are the most frequently used forms of non-personal communication for adult Americans, according to a new Gallup poll of communications behavior.

Between 37 percent and 39 percent of all U.S. residents said they used each of these “a lot” on the day prior to being interviewed. Perhaps the most-significant finding, however, is that just nine percent of respondents use a home landline phone over the same time frame.

About 15 percent of respondents reported using a landline phone at work the prior day.

All of that leads some to speculate that within a few decades, nobody uses a landline phone anymore.

That might be too extreme a prediction. As service providers already have discovered, bundling voice as part of a triple-play offer props up buying of voice, increasingly the least-in-demand service within the fixed network bundle. But the long-term trend is clear enough, since mobile phones now are the preferred way most people use voice communications.

So what comes after mobile Internet access, the present growth driver for mobile services, or video entertainment for fixed network telcos? What comes next for cable TV companies?

In addition to market share shifts, where cable TV providers have taken voice services revenue and more recently business customer “voice and data” share and telcos have been taking linear video share, the challenge of truly-new services remains.

In a broad sense, mobile and fixed network telcos and cable TV companies will grow by gaining scale across the whole business (mobile and fixed; domestic and international).

Beyond scale–across fixed and mobile domains–service providers are looking to Internet of Things or machine-to-machine opportunities, including vehicle communications, industrial sensor applications and potential consumer appliance connections (tablets connected to the mobile network are an immediate case in point).

Though there still is some work on creating applications that compete directly with over the top applications, the likelihood is that more money will be made supporting communications for billions of new devices, as an extension of today’s main revenue driver, Internet access.

Telecom executives are relatively unperturbed about dwindling voice revenues in large part because other replacement revenues already have been identified, while the next generation of new services and apps already are within sight. It will be hard work. But the roadmap is in place.

by Gary Kim

More Millimeter Wave Spectrum for Mobile Operators?

A frequent prediction about mobile bandwidth demand is that 1,000 times more capacity will be needed. How that can happen is the issue.

A rough list of solutions suggests that about 10 times more capacity will be gained by new bandwidth allocations, another 10 times increase will be gained by changes in network architecture and a final 10 times improvement will be gotten from applying better signal processing, using better antenna solutions and other operational efficiencies.

So the Federal Communications Commission now is looking at whether 24 GHz spectrum can be released for mobile communications applications, part of an overall effort to free up more licensed, unlicensed and shared spectrum for communications applications.

The Notice of Inquiry occurs at the same time that Google has asked for permission to test communications across different high-frequency spectrum bands, including millimeter-wave systems operating in the 71 GHz to 76 GHz band and the 81 GHz to 86 GHz range.

Ironically, as networks get faster, consumers respond by consuming lots more data, as well.

Long Term Evolution 4G networks have a rather predictable impact on mobile data consumption: the amount of data consumed each month grows, compared to data consumption on 3G networks.

In fact, some studies also suggest that access to LTE networks also increases use of Wi-Fi.

A study of Android smartphone users by Devicescape, conducted over six months, found that 4G LTE users consumption of Wi-Fi and mobile data doubled, compared to data consumed by 3G users.

On average, 4G smartphone users consume 2.1 times more mobile data per month and twice as much Wi-Fi than their 3G counterparts.

This is due to the fact that 4G customers use their mobile device about 40 percent to 50 percent more than 3G users and consume richer content. Also, as a practical matter, one minute of use of LTE results in more consumption than one minute of 3G usage simply because more data can be transferred in the same amount of time.

A September 2014 report by Citrix found that video accounted for 52 percent of all mobile traffic, on both 4G and 3G networks.

But 4G users were 1.5 times as many requests for video over 4G LTE networks than on 3G networks, and those requests resulted in five times as much video data traffic on 4G compared with 3G.

Ironically, the more the supply, the more the demand.

But advances in computing capabilities now make possible more extensive use of spectrum that in the past has been unusable for communications purposes.

“Years ago, engineers and policymakers debated the feasibility and practicality of using spectrum above 2 GHz for mobile wireless services,” FCC Chairman Tom Wheeler noted.

More recently, 3 GHz has been seen as the highest frequency that could be used to support mobile operations.

The difference now is signal processing that allows practical communications at frequencies traditionally unusable. But cheaper signal processing now means it is possible to overcome propagation issues that have prevented use of millimeter waves for mobile or fixed communications apps.

So there now is optimism that frequencies above 24 GHz could be used to support mobile service, a previously-unthinkable option.

This matters for obvious reasons. More spectrum is needed. Also, the basic trade off–capacity and distance are inversely related–means very-high capacity is possible at millimeter wave frequencies, even if distance is limited.

Physics dictates the higher bandwidth possible at millimeter wave frequencies, even if coverage is more limited than at frequencies below 2GHz. Despite digital coding, potential bandwidth still is dictated by the number of oscillations a radio wave makes in a single unit of time.

In other words, at the peak of the cycle, coders might represent a positive bit, at the trough, a negative bit. So the total number of possible symbols depends on the frequency, or number of instances in a given unit of time that the waveform crosses between high and low states.

As the name implies, higher frequency signals have many more oscillations than lower frequency signals. Hence, more potential bandwidth, using any particular coding and modulation scheme.

The trade off is the effective distances at which such waves are useful for mobile or fixed communications, as millimeter waves are attenuated by water and, in some cases, oxygen. The trick is to use frequencies where attenuation is relatively lower, as is the case for optical communications as well.

Still, it seems highly probable that new frequencies, best suited for use in dense population areas, will be released for service, at some point.

by Gary Kim