Early Millimeter Deployment Might Remind You of Early 1G

This map of Verizon’s 5G deployment in downtown Denver illustrates the use of 5G as an overlay on 4G for capacity reinforcement. The map also illustrates a feature of millimeter wave spectrum when used to support 5G or other networks: it does not propagate as well as mid-band or low-band frequencies.

Note the coverage basically outdoors, along streets and sidewalks. For some of you, that might harken back to the early days of analog, first-generation cellular service, which essentially was a "call from your car" service. That being the case, it made sense to build first along major streets and highways.

Note that 4G coverage is shown as pervasive. That illustrates the importance of non-millimeter wave frequencies to supply coverage, especially indoors. 

source: Verizon

How Much Demand for 5G in Germany, When 3G is As Poplular as 4G?

Based on past behavior, it is hard to forecast the strength of consumer demand for 5G in Germany. So far, nearly half of all subscriptions might use 3G, instead of 4G. nearly a decade after 4G was introduced.

Governments and policymakers always are quick to quantify gaps in uptake, quality or availability of communications services, which is among the reasons stories about any form of digital divide are evergreen. Most often, studies about service gaps rely on supply or demand indices, including network availability, typical speeds and cost. 

Demand side choices by consumers tend to overlooked. In other words, some “gaps” might reflect consumer choices, not failures of supply. And that matters for 5G, as much as it did for 4G. Consider the case of Germany. 

We often are surprised at the resilience of legacy services, as those use of legacy services is always a case of supply failure. Not always. Sometimes demand choices are at work. 

That appears to be the case for German mobile users and 3G. According to a recent study by Opensignal, as many as half of German subscriber identity modules (mobile accounts) are not enabled for 4G service. 

In other words, a huge percentage of German mobile users seem to be opting to remain on 3G networks even when 4G networks now are in good supply, with good to improving performance metrics. 

In other words, the question is why so many German users opt to remain on 3G networks when 4G networks are widely available and offer better performance. Low consumption of mobile data might be part of the explanation, some could argue. 

source: Statista

Compared to customers in other developed nations, Germans seem to consume less data. Some might point to high cost per gigabyte as a contributing factor. At least in 2016, German per-gigabyte costs were on the high side, compared to other European nations. 

source: Point Topic

But per-gigabyte costs in Germany have fallen since 2016, as they have virtually everywhere else in the world. 

source: Statista

To be sure, the actual cost of using mobile data depends on which plans users have. And international comparisons require choices about which plans to compare. That can distort results if the plans most-often used in particular countries are not uniform. In other words, if most consumers in some countries buy large-usage plans, others buy small-allocation plans, while in yet other instances most consumers buy on group or bundled service plans, not stand-alone mobile plans, comparisons are going to skew in ways that do not actually reflect buyer choices across countries. 

Consider this analysis of mobile internet plans featuring big usage allowances of 20 to 42 gigabytes per month, compared to the analysis above of one-GB usage plans. Where smaller usage plans might feature per-GB costs as high as $12 per gigabyte, when bought in higher volume (22 Gbytes in this case), cost drops to about $2.50 per GB. 

source: Babajob

Still, the point is that cost per gigabyte, or absolute cost, might not explain why so many German consumers choose 3G over 4G. 

Still, according to Opensignal, perhaps 81 percent of 4G non-users have for some reason elected not to buy 4G service. Some might suggest the “poor” state of German 4G networks explains why people do not buy. That seems unlikely. 

In a July 2019 study by Tutela, 4G downstream speeds averaged 23.5 Mbps on Telekom’s network, 21 Mbps on Vodafone and 18 Mbps on O2 networks. 

To be sure, German downstream speeds tend to be slower than in Switzerland or Austria, as measured by Tutela, but average downstream speeds--based on both 3G and 4G activity--of about 14 Mbps do not seem out of line on a global basis. 

Also, recent tests show typical 4G speeds in German cities ranging from about 25 Mbps to 35 Mbps, which does not seem slow, as far as 4G goes. 

source: Opensignal

Nor do German 4G speeds seem to have lagged, by about 2017, in comparison to other developed nations. 

source: Akamai

The point is that it is not self evident that poor 4G experience explains the lag in uptake. But some might point to coverage, rather than speed, as an explanation. 

“Whereas network providers in the Netherlands, Belgium and Switzerland can almost all offer LTE (4G) connections nearly 90 percent of the time, German communications giant Deutsche Telekom in Germany achieves only a 75 percent rate,” a study by consulting firm P3 says. Vodafone in Germany has 57 percent LTE coverage, P3 says. 

By some estimates, 4G coverage in Germany does not seem underdeveloped, though, and less coverage in eastern regions might be explained by greater rural character in the east. 

source: nperf

According to Germany’s Federal Network Agency (Bundesnetzagentur),  at the end of 2018, there were 107.5m active SIM cards in Germany (excluding M2M and IoT cards), but only 50.5m 4G/LTE SIM cards in active use. This would indicate that roughly half of the active SIM cards were not LTE-enabled

According to Opensignal, 81.4 percent of users that have never connected to 4G had a 4G-capable phone and spent time in 4G-covered areas. “These users likely did not upgrade to a 4G subscription or have disabled 4G connections on their phones,” says Opensignal.

Perhaps 15.6 percent of users who did not connect to 4G networks spent time in 4G-covered areas but did not have a 4G-capable device. 

A small percentage of non-4G users (perhaps three percent) appear not to live in rural areas where 4G is available. 

The point is that consumers sometimes choose not to buy the better or best mobile data plans, based on speed, price, availability (coverage) or other considerations. German consumers seem to be content buying 3G, instead of 4G, for some combination of reasons. 

So one wonders what the reception for 5G might be.

In the Near Term, Almost No Prediction about 5G Matters

Almost everything one can predict about 5G in 2020 must be considered transitory and conditional. Most basically, especially in larger countries, a new mobile network takes time to build, so coverage, subscription numbers, use cases, revenue impact and so forth will take some time to develop. 

Almost no claims made in the short term will matter longer term. That applies to fears about business models or hopes for new use cases. It is way too early to say anything definitive about potential innovations in retail models, revenue or “leadership” and “impact” in general, and 2020 will not change that.

The simple fact is that it might take five years before half of people in larger countries, or some mid-size countries, will be able to buy 5G service. So any conclusions about 5G impact will be made on a base of about half the population, most of whom will have had a chance to buy only in the latter two or three years. 

In other words, most customers will have been through about one handset cycle, after 5 years of 5G availability. It will be difficult to make long-lasting claims about behavior and results after just one handset cycle.  

Consider 4G, which launched by Verizon in 2010, with a buildout that essentially reached national coverage similar to 3G in about four years. The other national carriers took longer to reach similar coverage levels. 

source: FCC

5G coverage will be incomplete; even spotty. Older networks will continue to represent the primary connections (either 3G or 4G). 

Though small city states can deploy a whole new network rather quickly, a continent-sized network always requires five years or more to reach significant coverage of the population, and longer to achieve anything close to 98 percent coverage of the surface area. 

Since most people live in cities in most countries, population coverage and territory coverage are not synonymous. Still, U.S. 4G coverage reached about 55 percent of the population after about five years. In Europe, 4G coverage reached about 34 percent of people after five years. In East Asia, 4G reached perhaps 24 percent of people after five years. 

source: GSMA

The point is that it is reasonable to expect 5G deployments will reach about half of people in perhaps five years, at best, in North America, Europe or East Asia, with faster adoption possible in smaller countries. 

The point is that we are in for a period of at least half a decade where nearly every claim or assertion is only temporarily correct. Almost every numerical measure will change, continuously, as investment continues. 

Virtually the same might be said for innovations and changes in use cases, applications, subscriber numbers, pricing and other terms of service changes, revenue and cost structures, handset availability, handset cost and features. 

Early on, 5G will often be a capacity reinforcement strategy, as 4G was, early on. Marketing will be challenging, as always when a new network is coming online, as mass media casts too broad a footprint, stimulating demand where supply is not available. 

It is conceivable that some mobile operators will experiment with new retail packaging plans, perhaps moving to emulate fixed network differentiation based on speed tiers. 

In some markets where millimeter wave spectrum is available, some operators might expand unlimited usage plans or use millimeter wave spectrum to compete directly with fixed networks for internet access accounts. Up to this point, mobile spectrum has cost an order of magnitude more than fixed access, making it a poor substitute for fixed network services.

In some cases, millimeter wave assets will allow full competition for the first time. But it will take years before we are able to assess the magnitude of such opportunities, as network coverage and marketing will be limited for some years to come. 

As always, the next-generation 5G network will, at least temporarily, widen the gap between urban and rural services (speeds, especially). That always happens because new networks are built first in cities, where the payback is quickest and demand the greatest. 

“All of the above” will be a key feature of 5G spectrum usage, with different countries using different frequencies.

The bottom line is that any 5G trends will be tentative, possibly short-lived, and ever-changing, for the better part of a decade. That will be true for network coverage, applications and use cases as well as devices and revenue models. 

Mobile Device Costs a Barrier to Usage

People in some emerging markets who say they do not use mobile phones also indicate that device costs are a big barrier. Also, device costs are a major barrier for people who say they share a phone, a study by Pew Research finds. 

“When asked for the primary reason they share a mobile phone, an eight-country median of 34 percent say they do so largely because they cannot afford their own,” Pew Research says. 

A median of seven percent of people across 11 countries surveyed say they do not own a phone but do use someone else’s regularly. 

source: Pew Research

Sharing tends to be more common among adults with less education and lower levels of income, the report suggests. For example, 12 percent of Filipinos with lower levels of educational attainment report sharing a phone, compared with four percent of those with more education. 

Lower income is a factor in countries like India, where individuals with lower incomes (17 percent) are nearly twice as likely to share a phone as those with higher incomes (nine percent).

In India, significantly more women (20 percent) than men (five percent) report sharing a device with someone else. In other countries, there are small or no differences between men and women with regard to their likelihood of sharing a phone.

To be sure, respondents also indicate that the cost of mobile data at least occasionally find mobile data costs an issue. 

source: Pew Research

source: Pew Research

What is "Voice" These Days?

These days, when people talk about “voice,” they increasingly mean the use of voice assistants, the spoken interface to a computing device, not “talking on the phone.” To be sure, there are key differences: one is a service, the other a feature; one a conversation between people, the other a query method for apps and devices. 

source: Capgemini

That growing use of voice assistants has ironic implications, as when people talk about “using voice” for customer service. In some cases, entire interactions might be completed using only an assistant. In other cases voice assistants handle routine interactions, with interactions escalated to a live agent when needed. 

source: Capgemini

source: Capgemini

U.S. Streaming Video Customers Consume 520 GBytes Per Month

OpenVault now reports that U.S. consumers of streaming video now consume more than half a terabyte of data per month. Average consumption by cord-cutter subscribers was 520.8 GB, an increase of seven percent in the third quarter of 2019 alone.

Average U.S. subscriber usage on fixed networks was about 275.1 GB in the third quarter, a year-over-year increase of 21 percent.

During the same period, the median monthly weighted average usage increased nearly 25 percent from 118.2 GB to 147.4 GB, indicating that consumption is increasing across the market as a whole.

Some believe increased usage is a business opportunity for retail internet access providers. Others are not so sure. Usage once mattered greatly for telecom service provider revenues, as most revenue (and most of the profit) was generated by products billed “by the unit.” 

Capital investment was fairly easy to model, and profits from incremental investment likewise were easy to predict.

All that has changed, as usage (data consumption) of communications networks is not related in a linear way to revenue or profit, all observers will acknowledge. And that fact has huge implications for business models, as virtually all communication networks are recast as video transport and delivery networks, whatever other media types are carried.

Something on the order of 75 percent of total mobile network traffic in 2021, Cisco predicts. Globally, IP video traffic will be 82 percent of all consumer internet traffic by 2021, up from 73 percent in 2016, Cisco also says.

The basic problem is that entertainment video generates the absolute lowest revenue per bit, and entertainment video will dominate usage on all networks. Conversely while all narrowband services generate the highest revenue per bit, the “value” of narrowband services, expressed as retail price per bit, keeps falling, and usage is declining, in mature markets.

source: Sagheian

Some even argue that cost per bit exceeds revenue per bit, a long term recipe for failure. That has been cited as a key problem for emerging market mobile providers, where retail prices per megabyte must be quite low, requiring cost per bit levels of perhaps 0.01 cents per megabyte.

Of course, we have to avoid thinking in a linear way. Better technology, new architectures, huge new increases in mobile spectrum, shared spectrum, dynamic use of licensed spectrum and unlicensed spectrum all will change revenue per bit metrics.

source: Tellabs

Yet others argue that revenue per application now is what counts, not revenue per bit  or cost per bit. In other words, as for products sold in a grocery store, each particular product might have a different profit margin on sales, and what matters really is overall sales, and overall profit levels, not the specific profit levels of products sold.

So the basic business problem for network service providers is that their networks now must be built to support low revenue per bit services. That has key implications for the amount of capital that can be spent on networks to support the expected number of customers, average revenue per account and the amount of stranded assets.

source: Chetan Sharma

Why Revenue per Bit Falls Toward Zero

The fateful decision to build all global telecom networks on internet protocol, creating multipurpose networks, essentially means every network now has to be dimensioned (in terms of capacity) to carry the most-bandwidth intensive, low revenue-per-bit service, namely entertainment video, almost always a service that the access provider does not own, and therefore derives no direct revenue from supplying. And such video now dominates data volume on global networks.

That is but one reason why capacity prices tend, over time, to fall towards zero. Essentially, consumer service business models require low prices. The salient example are internet access services, where the internet service provider does not own the actual video services being watched. 

In the U.S. market, for example, consumers might use 300 Gbytes a month, with monthly revenue being perhaps $50, implying gigabyte revenue of 16 cents, and less if consumption is higher. 

Even if the access provider owns a subscription video service, it has the absolute lowest revenue per bit of any other owned service,  and therefore potential profit per bit, of any traffic type. 

Some might argue that an owned subscription video service has revenue per bit two orders of magnitude (100 times) less than voice, for example, in part because voice and text messaging use such little bandwidth, compared to video.

Text messaging has in the past had the highest revenue per bit, followed by voice services. More recently, as text messaging prices have collapsed, voice probably has the highest revenue per bit.

Subscription video always has had low revenue per bit, in large part because, as a media type, it requires so much bandwidth, while revenue is capped by consumer willingness to pay. Assume the average TV viewer has the screen turned on for five hours a day.

That works out to 150 hours a month. Assume an hour of standard definition video streaming (or broadcasting, in the analog world) consumes about one gigabyte per hour. That represents, for one person, consumption of perhaps 150 Gbytes. Assume overall household consumption of 200 Gbytes, and a monthly data cost of $50 per month.

Bump quality levels up to high definition and one easily can double the bandwidth consumption, up to perhaps 300 GB.  

That suggests a “cost”--to watch 150 hours of video--of about 33 cents per gigabyte, with retail price in the dollars per gigabyte range. 

Voice is something else. Assume a mobile or fixed line account represents about 350 minutes a month of usage. Assume the monthly recurring cost of having voice features on a mobile phone is about $20.

Assume data consumption for 350 minutes (5.8 hours a month) is about 21 MB per hour, or roughly 122 MB per month. That implies revenue of about $164 per consumed gigabyte. 

The point is that implied revenue per bit varies tremendously, even if networks now are sized to handle video, not the other media types. 

Voice and messaging arguably have the highest revenue per bit profiles (perhaps as high as hundreds of dollars per gigabyte). Both are applications sold by the mobile access provider and both consumer very little bandwidth. 

Entertainment video subscriptions sold by access providers do generate app revenue for providers, but relatively little, on a revenue per bit basis, compared to the narrowband voice and messaging services. Lots of bandwidth is required, but revenue is flat rate, so the actual revenue per bit hinges on usage. 

Then there are the “bandwidth” products supporting internet access, where the access provider generally earns zero money from app use, and rings the register only on gigabytes purchased. 

Entertainment video arguably generates cents per gigabyte in subscription revenue. That is a business model problem, as retail prices are in the dollars per gigabyte range. 

Mobile bandwidth  generally costs $5 to $8 per per gigabyte (retail prices), lower than the $9 to $10 it cost in 2016 or so, for popular plans, and less than that for plans containing higher amounts of usage. Higher usage plans might feature costs per gigabyte closer to $3.

The big point is that there is a reason why bandwidth prices tend to fall towards zero: consumer willingness and ability to pay for services and apps using bandwidth is relatively low, and does not change much over time.