Monday, August 31, 2020

Covid Impact Overstated?

An obvious storyline since March 2020 has been “impact of Covid-19 on X.” The problem with that story is that it might not be accurate. Looking at telecom service provider and infrastructure revenue growth trends, for example, and knowing nothing else, one would not necessarily conclude that anything unusual had happened. 


The reason is simply that, between 2019 and 2020, revenue growth already had been slipping. 

source: TBR


And that trend preexisted 2019. Looking only at telecom service providers in Western Europe between 2008 and 2014, revenue has dipped since about 2011. Covid-19 had nothing to do with those trends. 


source: STL Partners


Similar trends are seen elsewhere, showing low telecom service provider revenue growth rates. STL Partners believes global service provider revenue growth will average less than one percent per year through 2022, for example. 


source: IDATE


The point is that we might attribute too much influence to Covid-19 as a driver of near term trends. The long term influence might be even less meaningful, though that is not the conventional wisdom.


Sunday, August 30, 2020

Why U.S. 5G "Lags," and Why it Will Not Matter

One of the most-recurring stories about U.S. communications infrastructure deployment, app use or performance is that it “lags” what other countries achieve, especially in the early days of deployment. But even long-term indices show “lagging” U.S. performance. There’s a good reason for those trends. 


The bottom line is that it is quite typical for U.S. performance for almost any important new infrastructure-related technology to lag other nations. It never matters, in the end. 


Eventually, the U.S. ranks somewhere between 10th and 20th on any given measure of technology adoption. That has been the pattern since the time of analog voice. 


We often forget that six percent of the U.S. landmass is where most people live. About 94 percent of the land mass is unpopulated or lightly populated. And rural areas present the greatest challenge for deployment of communications facilities, or use of apps that require such facilities. 


For example, a recent study of 5G speeds by Opensignal found U.S. and U.K. speeds lagging those of several other countries that have deployed 5G. Some of the disparity in the U.S. market is based on the spectrum assets used to deploy 5G (low band and millimeter wave) rather than mid-band spectrum. 

source: Opensignal


But the “U.S. lags” story is quite typical. In fact, the “U.S. is behind” meme never goes away, where it comes to communications. The latest assertion in some quarters is that the United States is falling behind in 5G. That claim has been made many times about other key technologies and always has proven wrong.


In the past, it has been argued that the United States was behind, or falling behind, for use of mobile phones, smartphones, text messaging, broadband coverage, fiber to home, broadband speed or broadband price.


Consider voice adoption, where the best the United States ever ranked was about 15th globally, for teledensity (people provided with phone service). A couple of thoughts are worth keeping in mind. First, large countries always move slower than small countries or city-states, simply because construction of networks takes time and lots of capital. 


With the caveat that some rural and isolated locations never got fixed network phone service, not many would seriously argue that the supply or use of fixed network voice was an issue of any serious importance for the nation as a whole, though it is an issue for rural residents who cannot buy it.


Some even have argued the United States was falling behind in spectrum auctions. That seems almost silly given the amount of spectrum not being released for use in the U.S. market, roughly an order of magnitude more spectrum than previously available for mobile services. 


What such observations often miss is a highly dynamic environment, where apparently lagging metrics quickly are closed.


To be sure, adoption rates have sometimes lagged other regions. 


Some assertions are repeated so often they seem true. Among such statements are beliefs that U.S. internet access is slow and expensive, or that internet service providers have not managed to make gigabit speeds available on a widespread basis. In fact, gigabit coverage is about 80 percent, but take rates might be as low as two percent. 


Other statements, such as the claim that U.S. internet access prices or mobile prices are high, are not made in context, or qualified and adjusted for currency, local prices and incomes or other relevant inputs, including the comparison methodology itself. 


Both U.S. fixed network internet prices and U.S. mobile costs have dropped since 2000, for example. 


The point is that the United States never leads in infrastructure adoption or performance, especially in the early days of deployment. But even at full deployment, U.S. metrics tend to place the country somewhere between 10th and 20th globally. That has been true since the days when the only thing we measured was use of analog phone lines.


Did Covid-19 Change Martec's Law?

There is wide agreement that the Covid-19 pandemic has caused many technology adoption curves to get a temporary bump up in adoption, with growth then continuing on the curve already in place before the pandemic and its organizational response.  That is illustrated by the impact of the “cataclysmic event” on an underlying rate of organizational change. 

source: chiefmartec


In other words, firms and organizations are said to have experienced “a year’s worth of change in a month.” 


Martec’s Law was coined in 2013 by Scott Brinker, Hubspot VP. Martec’s Law states that technology changes linearly, while technological change is non-linear. That observation has parallels in the notion of the productivity paradox. 


The productivity paradox suggests that information technology or communications investments do not always immediately translate into effective productivity results. Many note that measured productivity has declined since 2000, despite all the technology investments firms have made. 


source: Goldman Sachs


This productivity paradox was apparent for much of the 1980s and 1990s, when one might have struggled to identify clear evidence of productivity gains from a rather massive investment in information technology.


Some would say the uncertainty covers a wider span of time, dating back to the 1970s and including even the “Internet” years from 2000 to the present.


The point is that it has in the past taken as long as 15 years for technology investments to produce measurable gains


Computing power in the U.S. economy increased by more than two orders of magnitude between 1970 and 1990, for example, yet productivity, especially in the service sector, stagnated).


And though it seems counter-intuitive, even the Internet has not clearly affected economy-wide productivity. Some might argue that is because we are not measuring properly. It is hard to assign a value to activities that have no incremental cost, such as listening to a streamed song instead of buying a compact disc. It might also be argued that benefits accrue, but only over longer periods of time


source: Customer Think


Few, if any, buyers of new technology actually believe the claims of benefit advanced by suppliers, for good reason. Virtually all observers of technology adoption note that organizations benefit from new technology at a rate that is vastly less than the rate of adoption. That’s the essence of Martec’s Law, which holds even if the Covid-19 pandemic caused an unusual step change in behavior. 


Saturday, August 29, 2020

What was Covid-19 Impact on Networking Services and Computing as a Service?

Two misconceptions sometimes abound about the impact of the Covid-19 pandemic on connectivity and computing suppliers. The first is that Covid disrupted operations and performance. To be sure, average end user experienced speeds dipped, in some countries, at least temporarily. But the big story is the one we did not hear: the networks did not crash. 


The other possible misconception is that the widespread business, school and organization shutdowns “must” have driven more revenue, given the centrality of work from home, entertain from home or learn from home activities. That turns out not to be true, either. 

source: Technology Business Research


Whatever gains might have been made from consumer customers, that was more than offset by the cessation of business activity by businesses large and small.


Friday, August 28, 2020

Can "High Tech" Displace "High Touch," or Does Virtual Drive Face-to-Face?

It should be obvious that what matters for businesses are outcomes, not inputs and activities: the financial bottom line (despite other stakeholder interests). Jobs, highly-desired quality products, sustainable environmental operations, employee benefits, contributions to economic vitality and other desirable outcomes cannot be obtained unless firms are sustainable financially. 


Also, some firms have a business model that benefits from remote work. Cloud services providers, PC manufacturers, connectivity companies generally and software firms whose products enhance remote work provide examples. 


So maybe we should not be surprised that Dell Technologies argues “COVID-19 has made one thing clear to us: work is something you do, an outcome, not a place or a time,” says Jeff Clarke, Dell Technologies COO. 


While the pandemic didn't start the remote work trend--it has been growing for 40 years, Covid-19 has shown its value, and almost certainly accelerates the trend. 


“Here at Dell, we expect, on an ongoing basis, that 60 percent of our workforce will stay remote or have a hybrid schedule where they work from home mostly and come into the office one or two days a week,” says Clarke. 


Such proclamations tend to convince observers that demand for other products, ranging from airline tickets to hotel stays, must inevitably suffer. And, to be sure, some period of lower demand likely will occur. 


But high tech drives high touch. virtual also drives more demand for face-to-face interactions.  That is a 40-year trend. This perhaps matters most for business-to-business sales professionals, who must establish trust with potential new clients. Eventually, when there is no medical bar to doing so, a renewed push for face-to-face contact will reassert itself, even if--and because of--high tech virtual replacements. 


U.S. Cable Operators Have 70% Share of the Installed Base

U.S. cable TV companies have about 70 percent share of the installed base of fixed network internet access lines, according to Leichtman Research. That might have seemed a shocking development three decades ago, when legacy telecom firms might have been assumed to have the inside track to lead the market. 


Virtually all observers would agree that this market structure exists not because “telcos cannot market” their services but principally because the cost of upgrading telco fixed networks to compete has a next-to-impossible business case. Simply put, the financial return from upgrading to fiber-to-home facilities is not there. 


That accounts for the interest in lower-cost alternatives such as 5G or 4G fixed wireless, 4G or 5G mobile access. 


Even with years of marketing, Verizon’s FiOS fiber to home network seems to get sustained market share of only about 30 percent. Across a base of 16 million homes, some note that Verizon seems essentially stuck at about that level of adoption.


AT&T has about 14 million to 15 million homes able to buy FTTH service. But AT&T seems relatively stable at about 30 percent share. 


CenturyLink fares worse, with FTTH take rates at about 11 percent to 17 percent. 


Those are daunting statistics, as they suggest--even when FTTH is available, most consumers do not buy it. In the United Kingdom, Ofcom estimates that 30 percent of consumers who could buy a fiber to home actually do so. 


AT&T believes it will hit 50 percent take rates for its fiber to home service after about three years of marketing, but so far those hopes seem unrealized. 


In South Korea, FTTH take rates seem to be only about 10 percent, for example, though network coverage is about 99 percent. 


In Japan and New Zealand, take rates have reached the mid-40-percent range, and network coverage might be about 90 percent. But in France and the United Kingdom, FTTH adoption is in the low-20-percent range. 


Perhaps 10 percent of Australians buy the fastest service available, whether that speed is 100 Mbps or a gigabit.

Thursday, August 27, 2020

Did Any Telecom Firms Actually Benefit from Covid-19?

The most interesting story that might emerge from the Covid-19 pandemic is whether any telecom or connectivity firms actually benefited financially.

Economic downturns--whatever the cause--are not “good” for connectivity providers, even at a time when reliance on communications, mobility and cloud computing might lead one to think revenues “should” be increasing because of that reliance. The Great Cessation of 2020 is not different. 

Service providers have experienced subscriber losses. A revenue contraction has widely been expected, is expected and earnings reports are starting to show that has happened. 


Customer spending dropped because of the Great Recession of 2008.  And one can infer that from job losses from the internet bubble collapse, the Great Recession and the present Great Cessation. 


source: Nordea, Macrobond


The collapse of the internet bubble in 2001, the Great Recession of 2008 and the Great Cessation of 2020 all lead to a cessation of growth at the very least, temporarily lower revenue for most connectivity providers, and bankruptcy for firms in some hard-hit sectors. 


It is not actually hard to explain why. Excess capacity and insufficient demand doomed many firms in the internet bubble collapse. Simply, when one’s customers go out of business, demand drops. Nor are connectivity providers exempt from the reduced expenditures characteristic of all economic downturns. 

Some might casually think the Great Cessation was different. It happened at a time when reliance on communications and remote computing is higher than ever before, but was intensified by mandatory “stay at home” rules. Businesses and schools were forced to close, instantly changing demand patterns. 


Factories stopped producing. Offices and retail businesses were shuttered. Travel fell off a cliff. Employees who got paid to work at home might not have noticed the harshest effects, but many millions of workers simply lost their jobs. Many millions of small businesses ceased to exist. 


And keep in mind that usage is not revenue. As much as at-home internet usage climbed, revenue did not, as such services often are not directly usage based. In fact, mobility and fixed network services often feature what is essentially “unlimited” usage, given the large size of usage buckets. 


In the U.S. market, most mobile customers have effectively unlimited voice calling and text messaging, while fixed network voice and internet access also is effectively unlimited. Most large internet service providers additionally removed usage limits during the lockdown, so usage limits formally vanished. 


So the business impact was that sharply higher usage did not drive revenue higher. Bouygues Telecom, for example, reported improved financial performance “after the lockdown ended.” Some might be tempted to spin that as a “Covid helps” story, but that is wrong. Revenue did not grow until after the lockdown was ended. 


A similar story is almost certain to play out at other telcos. Recessions simply are not good for business.

Tuesday, August 25, 2020

Is Network Slicing a Feature or a Service?

Will network slicing, a new feature of virtualized 5G and other network cores, enable mobile operators to move into new parts of the value chain? Put another way, can service quality assurances made possible by network slicing do so? 


Or is network slicing a platform upon which to create such new roles? Some, including Netrounds, believe network slicing, almost by definition, moves mobile operators up the value chain, if not necessarily into new roles within the value chain.


That is the challenge. Is there enough demand for service level agreements to provide significant revenue-generating value? In the absence of other changes--taking on new roles in the ecosystem, for example--is QoS enough? Or is network slicing “necessary but not sufficient” for value creation?


In other words, is network-slicing-enabled QoS so valuable that customers will pay extra for it, and if so, how much? Is that move up the value chain--providing more value--enough? Or, in the end, must that also be coupled with moves into other parts of the ecosystem, such as applications and platforms, for example?


To use the personal computing analogy, are better “speeds and feeds” sufficient to drive revenue growth, or not?


To be sure, network slicing does offer some fundamentally new capabilities compared to previous generations of mobile networks. Up to a point, virtual private networks can be optimized around some relevant performance characteristics, end to end. 


source: Nokia


A mobile broadband network--possibly a business-to-business mobile virtual network operator--could be optimized for bandwidth. An autonomous vehicle network (aerial or terrestrial vehicles) might be optimized for latency. A sensor network might be optimized for low power consumption. 


Perhaps a gaming network is optimized for bandwidth, latency and jitter. Stadium and entertainment venues might be optimized for device density and bandwidth. 


source: SDX Central


In other cases, perhaps it is the billing method that is new: pay per use models, for example. Dynamic or temporary services are another potential use case, with at least some potential for incremental revenue. 


As always, there are challenges, not the least of which is that there always are different ways to meet each of those needs. Edge computing, for example, can provide the ultra-low-latency or bandwidth assurance features. Small cells can provide the device density features. 


Perhaps it is the on-demand, dynamic provisioning and rating that provides the biggest differentiator, if the cost of spinning up and spinning down services is low enough. 


Still, there will be other ways of creating customized networks built to support specific applications. Edge computing, in particular, might be a functional substitute for network slicing, for example.

Sunday, August 23, 2020

SpaceX Starlink Has Big Ambitions

The SpaceX Starlink home broadband service illustrates one important fact about new markets for internet service providers, namely that the amount of new revenue is more limited than one might think, at a time when a shift of just a few points of market share arguably has greater financial impact than most new revenue streams put together. 

Assume the U.S. consumer broadband market--mobile and fixed--is close to $200 billion in annual revenue by about the time Starlink is able to sell gigabit service. A shift of just one percent market share is $2 billion. If the market is mostly a zero-sum game, then when one competitor gains a point of share, the market change actually is two percent (plus one for the winner, negative one for the ISP that loses the share). 


For competitors in the U.S. home broadband market, the impact of losing one market share point is arguably greater than all new service revenues put together. In other words, in the short term, it is rational to defend legacy market share as “job one,” since that has the most telling impact on revenues and profits. 


Attackers have a different problem. Taking market share in an existing large market is a time-tested recipe for growth. 


SpaceX expected in 2017 that by 2022, Starlink revenue would account for roughly 75 percent of all SpaceX revenue. By 2023, Starlink satellite internet access was projected to represent more than 80 percent of total SpaceX revenue, reaching 85 percent by 2025.


If Starlink gets anywhere close to those numbers, it likely will be because Starlink has become a significant presence in the U.S. home broadband business, partly by connecting rural customers whose choices are restricted to satellite service, but also many rural customers who have access to cable TV, telco, fixed wireless or mobile internet access. 


Starlink believes it could get as many as five million U.S. households connected if it can put enough satellites into orbit, allowing it to boost speeds up to around a gigabit per second. At $80 per month, that could generate as much as $4.8 billion in annual revenue. 


That implies Starlink takes about 2.5 percent share of the U.S. home broadband business. Of course, revenue is one issue. The cost of launching the full constellation is the other key issue. Ability to deliver gigabit speeds might require something closer to 42,000, compared to the 800 or so satellites Starlink will need to provide commercial service at speeds up to 60 Mbps or so. 


The capital investment for the full constellation is three orders of magnitude larger than the initial constellation. 


Back to legacy providers: most of Starlink’s gains will come at their expense. In other words, mobile and fixed consumer internet providers will lost nearly $5 billion if Starlink gains that much. The global multi-access edge computing market might range between $1.5 billion and $5 billion by about 2025, when Starlink might hope to have increased its fleet closer to 42,000 satellites. 


source: STL Partners


The point is simple: incumbents have more to lose from lost home broadband market share than they have to gain from multi-access edge computing, and that likely is true even adding in all other new revenue streams such as the internet of things and private networks. If the global MEC business ranges between $1 billion and $5 billion, U.S. suppliers could not hope to gain more than $500 million to $2 billion of that revenue, even if the U.S. market represented half the global market. 


To be sure, it is not hard to find estimates of global connectivity revenues approaching $40 billion per year, across all platforms. And there are lots of alternatives for IoT connections. 


source: Markets and Markets


To be sure, a brand new revenue stream generating $1 billion or more is a wonderful thing. But those types of new businesses are hard to create. If attacking firms such as SpaceX and T-Mobile are able to achieve their stated objectives, defending firms looking to edge computing, internet of things and private networks for growth might find the new revenues only replacing what they lose to attackers in the home broadband business. 


Gross revenue from new sources is essential, for attackers and incumbents. Net revenue change might be underwhelming, in the near to medium term. 


Saturday, August 22, 2020

"High Tech, High Touch" Patterns Will Return

The prevailing wisdom about business life after Covid-19 often is that “nothing will be the same.” In place of “high touch” face-to-face meetings, businesses are going to substitute “high tech” virtual sales and marketing. That might happen, to a significant and permanent degree. 

At the same time, firms are going to rediscover the value of face-to-face, “high touch” activities, perhaps to the same degree as they shift to virtual “high tech” operations. It has been a 40-year trend. 


Some of us remember a 1982 book called Megatrends by futurist John Naisbitt that popularized the phrase “high tech, high touch” to describe a coming trend: that as we began to use more technology, we would equally appreciate “high touch” human or non-technological interactions. He explored how that had evolved in his follow-on book High Tech, High Touch


Here’s the important insight into post-Covid-19 business behavior: High-touch refers to the human and emotional aspects of business interactions, including the establishment of trust


Virtual support can be quite effective once a business relationship and trust have been established. But it arguably will be harder to create trust with a new prospect using only “high tech” tools. 


High-touch refers to close relationships with customers, interacting with people, not just machines.  To be sure, that arguably can be done, up to a point, virtually. But high-touch also arguably requires above-average interaction with customers, and that includes face-to-face contact. 


High touch--helping customers on a human level through various stages of the buying process and lifecycle--involves a much higher participation, and usually relies on one individual or team within the company to maintain direct, personal and frequent contact with accounts, says ESG. 

“Humans crave the kind of interaction that only other humans can provide,” and likely cannot always be provided by conferencing tools. 


It’s one that places the priority on human interaction and human relationships, not on efficiency or speed. That is why travel services emphasize both touch and tech. 


“In a high tech world, people are longing for balance,” notes EHL Insights. As important as technology has become for customer experience and support, “authenticity” and “emotion” also are emphasized. There is an analogy for business-to-business sales as well. 


This “reversion to mean” happens quite often. Right now, professionals now rely on videoconferencing to supplant face-to-face meetings because nothing else is possible. But after the pandemic ends, prior trends will reassert themselves.


That has proven to be true for industries and economies in recoveries from major economic disruptions in the recent past. 


If evidence from three past global recessions--but not the Great Recession of 2008 and the Great Cessation of 2020--provide any useful insight, the recovery might take between three and 4.5 years, perhaps three years for public companies, perhaps 4.5 years for the overall economy. 


Some 17 percent of public  firms will not survive. That is the percentage of public firms that went bankrupt, were acquired, or became private, in the aftermath of the Great Recession. 


About 80 percent of the survivors had not yet regained their pre-recession growth rates for sales and profits three years after a recession. 


About 40 percent of the firms had not returned to their absolute pre-recession sales and profits levels after three years. 


Ranjay Gulati, Harvard Business School professor and Nitin Nohria, Harvard Business School dean, conducted a study in 2010 of corporate performance during three global recessions: the 1980 crisis (which lasted from 1980 to 1982), the 1990 slowdown (1990 to 1991), and the 2000 bust (2000 to 2002). 


Obviously, the two big events missing from the study, because of the timing, were the Great Recession of 2008 and the current “Great Cessation” of 2020. Still, their findings are useful for charting the likely path of recovery after the Covid-19 pandemic recedes into history.


They studied 4,700 public companies, breaking down the data into three periods: the three years before a recession, the three years after, and the recession years themselves. 


At a macroeconomic level, the U.S. economy had not recovered its pre-2008 levels by 2011, three years after the Great Great Recession of 2008. 


source: Bureau of Economic Analysis


U.S. growth rates returned to 2007 levels about 4.5 years after the Great Recession. Latin America and some Asian countries bounced back really fast, in about 1.5 years.


So if the recovery from the Great Cessation follows the Great Recession pattern, it will take about four years for gross national product to return to 2019 levels. 


The impact on household wealth was starker, as median household net worth still had not reached 2007 levels by 2018, 10 years after the Great Recession. Job levels in the United States had returned to 2007 levels by 2014 (about six years after the Great Recession). 


Still, prior trends reasserted themselves. That is likely to happen with face-to-face “high touch” sales in the business-to-business markets as well. It might take some time, but it is almost certainly going to happen. 


"Tokens" are the New "FLOPS," "MIPS" or "Gbps"

Modern computing has some virtually-universal reference metrics. For Gemini 1.5 and other large language models, tokens are a basic measure...