Thursday, January 31, 2019

Retailers Prep for IoT

According to a new study by BRP Consulting, 19 percent of surveyed retailers now some form of Internet of Things (IoT). Another 36 percent of respondents plan to do so within three years.

Some 23 percent of survey respondents say they already use artificial intelligence and an additional 30 percent plan to do so within three years.


Some 32 percent of surveyed consumers say they are likely to shop at a store offering an augmented reality experience. Some nine percent of retailers said they would offer augmented reality to their customers and another 29 percent planned to do so within three years.
Some 29 percent of consumer respondents said they are likely to shop at a retailer offering virtual reality, while seven percent of retailers said they already do so. Some 23 percent said they plan to add it within three years.

Enabling this shift to information-based value creation are the falling costs and increasing functionality of sensors.  Sensor prices have tumbled to about 40 cents from $2 in 2006, with bandwidth costs a small fraction of those even five years ago, according to Deloitte researchers.  

Has AT&T Changed its Mind about 5G Fixed Wireless?

Over the last couple of years, AT&T has seemingly waned and waxed about commercial upside for 5G fixed wireless. For most of 2018, AT&T officials had expressed skepticism about the business model.

That thinking might have been based on use of millimeter wave frequencies. The ability to use spectrum in the 3.6-GHz region (Citizens Broadband Radio Service) should require less-dense cells and backhaul, and might also tip towards more business customer use cases, rather than consumer.

All that might have brought a change of views about use of fixed wireless.

“I will say over time three to five year time horizon unequivocally 5G will serve as a broadband, a fixed broadband replacement product,” says AT&T CEO Randall Stephenson. “I am very convinced that that will be the case.”

“You know, we back in the 90s everybody was saying wireless would never serve as a substitute for fixed line voice because there wasn't sufficient capacity,” Stephenson said. “Well it is a substitute for voice.”

Right now, AT&T says it has 11 million fiber to home locations and eight million business locations. AT&T also expects to reach 14 million consumer fiber-to-home locations soon. It probably is worth noting that AT&T’s fixed network passes--is able to sell services to--as many as 62 million U.S. homes.

In other words, AT&T might soon pass 22.5 percent of its consumer locations with optical fiber drops.

Even without quantifying the matter, if AT&T has managed to build optical fiber to less than a quarter of its U.S. homes, and also believes 5G will provide a workable substitute within three to five years, it is hard to see the logic of continuing to build consumer optical fiber connections, at a time when consumer fixed line accounts are shrinking overall.

Wednesday, January 30, 2019

Telenor Says 5G Will Not Boost Capex

Norway's Telenor, which plans to launch 5G services in Norway in 2020, also says it has no plans to boost overall capital investment (except for the costs of spectrum).

The operator said it would invest between 16-17 billion Norwegian kroner (US$1.9 to $2 billion) in capital expenditure in 2019, excluding the cost of spectrum licenses, after spending  NOK16.8 billion ($2 billion) in 2018.

That seems to be part of a trend in developed markets, where business risk is mitigated, in part, by phased deployment to complement 4G, which also is getting better. Also, new studies suggest 5G capex requirements will not be as onerous as originally believed.


More to the point, several mobile service providers already are saying that 5G capital expense will not exceed present levels of spending, or will increase only slightly, as is typical whenever a next-generation network starts to be deployed. Swisscom, 3UK, Verizon and AT&T are among firms in addition to Telenor signaling flat capex as 5G is introduced.  

That might come as quite a relief t some, a surprise to others, who had forecast 5G capital investment so high many operators could not afford it.  

Monday, January 28, 2019

Will 5G Be "Too Expensive?"

Some observers suggest (in advance of any concrete information) that 5G will be expensive. Some consumers worry about that as well. Some in the mobile industry might wish that was going to be a problem. But history suggests “high cost” is not going to be an issue for 5G as deployed in the U.S. market, or elsewhere, for that matter.

For starters, in most developed markets, mobile data access simply is not expensive.

Of course, one might make the argument that a bandwidth boost of as much as 1,000 times sometime in the 5G era will have consequences. To keep retail prices about where they are today, the implication is that costs per delivered bit would have to drop drastically by 1,000 times as well.


That ignores history. Since the start of the internet age, internet access bandwidth costs have dropped. That has been true for both fixed and mobile networks.

With massive amounts of new spectrum coming (millimeter wave bands alone will increase physical capacity at least 10 times. Massive use of small cells will increase effective capacity another 10 times. Spectrum aggregation, use of unlicensed spectrum, better radios and modulation techniques will increase effective bandwidth by perhaps another 10 times.

Add it all up and supplying three orders more bandwidth is possible, both technically and in terms of the existing business model, though not easy.



One might note that most countries cluster in a range of usage around four gigabytes per user SIM per month, and prices around $5 per gigabyte per month. The countries with higher consumption per user also tend to have lower-than-average prices, say researchers at Tefficient.

Granted, posted retail rates for mobile data can run about $10 per gigabyte, for some providers, on plans with lower usage allowances. For plans supporting bigger data allowances, costs are $5 per gigabyte or less, depending on plan and carrier.

And then there is the statistical reality. If a user pays for 30 GB, and uses just 5 GB a month, the effective data cost is far higher. Buying the right to use 30 GB is one thing. When actual usage is lower, the effective rate of “consumed” data is far higher.

Almost nobody actually thinks about data prices that way, though. Instead, people tend to compare only the sunk cost per month, so long as the plan covers the anticipated usage.

Prices less than $5 per gigabyte are viewed by some as “too high.” In the U.S. market, it is easy to find a supplier selling such usage at perhaps $5 to $7 per gigabyte, for just a few gigabytes per month (3 GB to 5G, for example), and bigger packages at lower prices per gigabyte.

With the caveat that developing nation prices might be higher, expressed as a percentage of income to buy the rights to use a gigabyte of mobile data, in developed markets, communications costs are pretty low, as a percentage of income.

The point is that even if there is some price premium early on for 5G mobile data, the premium is likely to be slight, and diminishing over time.

Nor, given the new spectrum resources, including huge amounts of millimeter wave spectrum, small cell  architectures, modulation techniques and radios, plus use of unlicensed, shared and aggregated spectrum, are suppliers going to need to raise prices by three orders of magnitude.

No, 5G prices will not be “too high.”

Sunday, January 27, 2019

Net Neutrality: Confusing Ends and Means

Telecom and internet regulators sometimes confuse “ideas” or “outcomes” with “policies” and “procedures.”

The problem is that desirable or intended outcomes cannot be mandated; only processes and rules that specify how an objective is to be achieved. And in a business as complex as telecommunications, policymaking too often turns on the desirable outcomes, not the policies to promote such outcomes.

At best, that is a simplistic or superficial way to set policy; at worst, a virtual guarantee the the expected or hoped-for results will not be reached.

Those were some of the themes Dr. George Ford, Phoenix Center chief economist, discussed from the center stage at the recent PTC’19 conference.

For example, “an ‘open Internet’ is an idea, not a policy,” said Ford. Open Internet is an outcome, not a set of policy procedures the Federal Communications Commission might create. So an open internet is to clean bathrooms as “no blocking of lawful content” rules are to checklists about when a bathroom is to be cleaned, what is to be cleaned and with what disinfectants.

“This confusion over ideas and policies is endemic to nearly all modern policy debates,” Ford says.

In the U.S. context, an “open internet” is one that permits “no blocking” of lawful content and “no fast lanes.” Those are examples of outcomes” (ideas, not policy).

Title II regulation is a policy originally developed to govern voice services. Known as “common carrier” regulation, Title II’s guidelines for setting prices is that they be “just and reasonable.”

Both objectives--no blocking and no fast lanes--actually violate Title II pricing rules. “No blocking” means zero price. And zero price, in a common carrier context, also means “zero cost.”

Though each instance of internet access use by any user does carry some non-zero cost, that cost is close to zero, but not zero. A “zero” price level is unlawful, in that case.

Second, “no fast lanes” implies that edge providers receiving different levels of service cannot be charged different prices, Ford says. But Title II “allows, if not mandates, different prices for things with different costs.”

The problem, Ford argues is that popular understanding was that the issue was the “openness of the Internet.” That never was the case. The issue was the merit of the Title II, common carrier policy as a way of preserving Internet openness, versus all other possible approaches.

Instead, one particular policy tool was said to represent the open internet; the rival positions as opposed to the open internet. That was a false and misleading choice.

Some might also argue that fast lanes already exist, in the form of content delivery networks. Also, internet access pricing is differential. Slower speeds cost less; faster speeds cost more. Usage limits likewise are differential. Neither affects internet openness.

In real life, we often say there is a difference between  ends and means. Ideas are outcomes or ends. Policies are means. The policy debate was not helped by confusing ends for means, or means for ends.

U.K. Household Communications Spending is Flat and Fixed

U.K. consumers, like consumers anywhere, are likely to spend only so much on communications services of all types. That explains why household spending on communications (including subscription video) is constant, year after year.

So when spending on some elements grows, spending on other elements has shrunk. In the United Kingdom, as video entertainment and fixed network internet access have grown, spending on mobile services and fixed voice have dropped.

That is worth keeping in mind when examining price levels (“the cost”) of internet access, mobile service, video subscriptions, messaging or voice. To buy more of something, all other things being equal, consumers will buy less of something else.

One problem we always have when trying to determine what people actually pay for various communications services is that we must work with “list” and “discounted” prices. It often is hard to tell what percentage of customers pay the list prices or some other discounted price (through a bundle or some other promotion).


The other issue is that prices tend to climb, over time, with inflation of general price levels. For example, the buying power of $100 in 1913 was $2564 in 2018, according to the U.S. Department of Labor. The buying power of $100 in 2008 was $119 in 2018.

So even if there were no improvements in quality, and all prices rose evenly, the $100 spent on any product in 2008 would be equivalent to spending $119 ten years later.

The consumer price index also tends to track growth of gross domestic product, over time.



The point is that, looking at price levels, it is natural for prices to increase over time. It never is equally clear whether prices are risking in “real,” inflation-adjusted terms. Nor is it easy to capture changes in demand or product quality.

I used to pay as much as $300 a month for a 768 kbps internet access service. Now the list price for service at 100 Mbps to 150 Mbps is perhaps $50 (before taxes and fees).

The important point is that on a cost-per-bits-per-second basis or even in absolute (not inflation adjusted)  terms, and not accounting for other discounts, perhaps most consumers pay the same, or less than they used to,  for fixed network internet access.

Nominal prices might rise over time, as prices generally do. But cost per bit and qualitative improvements are not to be captured in the retail price data.

SIP School Survey Shows 9% of Businesses Rely Exclusively on Hosted Voice

According to the latest survey conducted by The SIP School, including responses from 374 global end user executives, 48 percent of respondents work for firms that use IP private branch exchanges (business phone systems). Just nine percent report using a hosted voice system, while 35 percent use a mixture of premises phone systems and cloud telephony.

Less than four percent use a digital phone system.


And if the economics of hosted voice do not change (recurring cost), premises phone systems at larger enterprises might never reach very-high levels.

The reason is simply that although hosted voice total cost of ownership works for the small business and mid-sized organizations in many cases, total cost of ownership still favors the premise-based PBX for larger enterprises, if the actual useful life of the switch is more than five years, even for a call center deployment, where call features are more important than for a typical enterprise.

Small entities might find the hosted solution more affordable for up to 10 years. After that, even a small entity might have a lower total cost of ownership in year 11, by some estimates.


To be sure, small organizations represent the volume of business entities, if not the volume of “seats” or “users.” In principle, most businesses will use cloud voice at some point, if legacy digital voice services are shut down in favor of cloud or hosted voice.

Of perhaps 5.8 million U.S. companies, only about 19,076 have more than 500 employees. But those are the entities most likely to find premises solutions most affordable. The exceptions are those large enterprises with a substantial number of branch and smaller locations or highly-dynamic staffing needs.


Results such as these always surprise me, though I admit I do not follow enterprise voice trends (though I used to do so). Had you asked me nearly two decades ago whether business entities (small, medium or enterprise) would have switched mostly to hosted voice by 2020 , I’d have guessed something like 70 percent to 80 percent would have done so, at least in the U.S. market.

It does not appear that has happened. Also,  40 percent global adoption (at least in part) by businesses seems high, though 40 percent adoption might be low for the U.S. market. The SIP School data suggests that perhaps nine percent of respondents only use hosted voice, while 35 percent of hosted voice adopters use a mix of premises and cloud solutions.

Gartner has estimated adoption of cloud-based business voice by perhaps 24 percent of organizations by the end of 2018. That probably includes any mixed-use instance, and likely does not imply 100-percent reliance on cloud voice.

I would still have guessed that adoption in other markets would be different. As I recall, hosted voice adoption still is highest in the United States.

The latest SIP School survey has about 40 percent U.S. respondents, so it is very likely the U.S. hosted voice adoption percentages are higher than the global totals, in terms of how organizations source voice services.

That would be true because hosted voice cost of ownership works for the smaller organization, compared to use of a premises PBX, because small entities represent most businesses, and because U.S. hosted voice adoption historically has been higher than the global adoption rate.

Still, I have been surprised that hosted voice has not been adopted faster. But the economics of a hybrid approach simply remain compelling: premises switches for large locations; cloud for everything else.

Saturday, January 26, 2019

Brand Finance Global 500 Shows Dominance of Internet Ecosystem Firms

The latest Brand Finance Global 500 study might be an illustration of the power and influence of the broad internet ecosystem. You might note that firms with device, commerce, application or networking roles occupy the top nine of ten spots at the top of the listings.


Also, the growing role of commerce, advertising and other platforms is clear. “Change at the top is reflective of a wider global trend as the technology sector accounts for more than twice as much brand value as telecoms,” the latest study notes.

The other obvious trend is growing share of Chinese firms. “The growth of Chinese brands extends beyond the technology sector as the country continues to narrow the value gap with the United States at an impressive rate,” the report says. Since 2008, China’s share of global brand value has increased from three percent to 15 percent.

Among telecom firms, Telstra “continues to perform exceptionally well across a number of brand attributes, with a resulting 14 percent increase in brand value to US$12.4 billion,” the report states.

Still, the study notes that brand strength and market value are not synonymous. “The strength of the brand is in contrast to the overall decline in Telstra’s market value of 21 percent,” the study says.

So brand value does not always translate into equity value.

New Study Finds "No Significant Effect" of Broadband on Productivity

The consensus view on broadband access for business is that it leads to higher productivity. But a new study by Ireland’s Economic and Social Research Institute finds “small positive associations between broadband and firms’ productivity levels, none of these effects are statistically significant.”

“We also find no significant effect looking across all service sector firms taken together,” ESRI notes. “These results are consistent with those of other recent research that suggests the benefits of broadband for productivity depend heavily upon sectoral and firm characteristics rather than representing a generalised effect.”

“Overall, it seems that the benefits of broadband to particular local areas may vary substantially depending upon the sectoral mix of local firms and the availability of related inputs such as highly educated labour and appropriate management,” says ESRI.

In other words, broadband--in and of itself--might not be a general driver of enhanced productivity or economic growth, absent other more-basic drivers. Some earlier studies do not clearly establish a direct relationship between productivity and broadband, except in the computing and information technology industry itself.

To be sure, economic activity and broadband are correlated. What is harder to determine is the degree of any causality, though many studies suggest broadband is correlated with economic growth, productivity and income.   

ESRI does note that  “significant gains from broadband availability in two services sectors: information and communication services and administrative and support service activities.”

Some have looked at the application of information technology and called it a productivity paradox. “You can see the computer age everywhere but in the productivity statistics,” Professor Robert Solow famously said.

In fact, productivity sometimes has dropped after more information technology was deployed. Of course, there could be other explanations. Productivity might have dropped for other reasons. IT might have prevented faster productivity declines. Perhaps we can not measure the productivity gains (qualitative improvements perhaps cannot be quantified, by definition).   

Some believe the earlier productivity paradox in computing now is seen in cloud computing, the internet and mobility as well, though productivity did seem to rise in some industries. Perhaps few would argue that information technology does not matter.

Still, despite the instances of correlation, causal relationships remain tough to identify. We note that consumption of high-quality broadband services is correlated with high income and education levels, for example. We might note high correlation between broadband and economic activity. Still, a causal relationship never is completely clear.  

It might be the case that high economic activity attracts highly educated and better-compensated people, who then have the resources to buy quality broadband, and live where that can be accomplished.

"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...