Friday, February 27, 2015

U.S. ISP Profit Squueze Now is Likely

The decision on Feb. 26, 2015 by the Federal Communications Commission to preempt state laws prohibiting municipal broadband networks is but one example of growing competition in the U.S. high speed access market that arguably already was getting more competitive, thanks largely to Google Fiber.


Consumers will benefit, at least in the near term, as the actual presence of a third provider in a growing range of local markets both increases speeds and lowers prices.


Consider what Sonic.net is doing. It now is selling gigabit connections, with voice service, for $40 a month, submarining even Google Fiber pricing of $70 a month for gigabit high speed access.


Cumulatively, all the coming new competition, plus declines in mobile high speed access, are going to mean service providers must reduce their operating costs.


From 2010 to 2013, U.S. mobile data pricing (per unit sold) declined by only single digits year over year.


But in the first nine months of 2014, data pricing dropped by 77 percent, according to industry analyst Chetan Sharma.


At the same time, average (mean) mobile data consumption increased to about 2 gigabytes (Gb) a month. Sharma notes it took 20 years for consumption to reach 1 Gb per month usage levels.


The increase to 2 Gb took about a year.


In addition to plunging prices (less revenue per unit sold) and higher usage (more network cost), marketing costs have grown as competition has become more intense.


Overall U.S. operating expense rose 20 percent, year over year. Income was flat while earnings grew three percent.

Like it or not, more competition in Internet access is coming. Profits are going to get squeezed.

Will Customers Pay for Two High Speed Networks--on Different Platforms--to Support Dual Routers?

If you have lemons, make lemonade, an old adage suggests.

For a high speed access provider with access to both cable TV hybrid fiber coax and wholesale fiber-to-home access, that might mean trying to add value to an offer by bundling access to both networks as part of a single high speed access subscription.

It’s unusual, but few service provider anywhere in the world have such access as does StarHub in Singapore.

So Starhub now is testing the notion that having two high speed connections, supplied by different networks, has value for Wi-Fi coverage, and therefore makes the Starhub “two connections” offer distinct in the Singapore high speed access market.

Essentially, StarHub is trying to wring more value out of a legacy HFC access network, now that it also has wholesale access to the fiber-to-home network in Singapore. Bandwidth really isn’t the angle.

It easily can be argued that there is actually little advantage gained  when a StarHub subscriber buys a S$69 per month gigabit access service sourced by the Next Generation Network Broadband Network used by all high speed access providers in Singapore, and then also gets, at no extra charge, a 100 Mbps connection to the StarHub HFC network.

Competitors offer the gigabit connection for S$49 a month.

In truth, few, if any applications actually use much of a gigabit connection, so adding another 100 Mbps might provide little if any value, in terms of the access connection’s bandwidth. And the additional cost will be a barrier.

But StarHub already has built the HFC network, and now it essentially is a stranded asset. So StarHub might as well try and figure out a way to wring some value out of the legacy asset.

In what might otherwise be an obvious case of overkill (over-provisioning or over-investment), StarHub argues the value is multiple Wi-Fi base stations, to improve indoor coverage.

It remains to be seen whether that “advantage” actually resonates with customers. But it is an unusual offer.

To be sure, StarHub sells linear video entertainment, so the HFC network investment is not fully stranded. But the NBNBN also now means StarHub has to offer gigabit services over the Singapore wholesale network, to remain competitive in the high speed access business.

That means the financial return from the HFC network will be driven primarily by sale of voice and linear video services, not a triple play bundle including high speed access. But that’s where the new “dual-network” approach comes in.

Over the long term, it isn’t so clear that it makes sense to maintain the HFC network. In the meantime, StarHub will try to earn a potential S$20 a month in incremental revenue by supplying a “two Wi-Fi routers” approach.

When you have lemons, make lemonade. Then, can you sell enough of it to make the effort worthwhile?

Thursday, February 26, 2015

Net Neutrality Rules Will Pass Today

The unknowns are the precise details, and even those won't be available for a few weeks. The rules would not take effect until 30 days after publication. 

Then there will be a short stay of the rules, as lawsuits are filed. Then, in all likelihood, the rules will take effect until the lawsuits are heard and decided. 

How much uncertainty will be created in the U.S. telecommunications business is an issue. 


Despite Robust Mobile Revenue Growth, Axiata Profit Drops 80%

An eight-year robust revenue growth streak has ended abruptly at mobile services supplier Axiata. The issue now is whether the dip is cyclical or structural.

In many key respects, Axiata Group now faces business issues identical to those faced by other mobile service providers, namely robust growth in mobile data revenues, offset by losses of voice and text messaging revenue.

But slower economic growth, foreign exchange issues and investments played a role in Axiata 2014 results as well.

Not even robust mobile data revenue growth of 50 percent was enough to overcome business headwinds faced by Malaysia-based Axiata Group in 2014, which had a full-year 7.9 percent decline in net profit on a 1.9 percent growth in revenue, where the target had been 10 percent.

Axiata has an 84-percent dividend payout ratio, so any dip in revenue has key implications for profits.

Total data revenue grew 31.5 percent year over year.

Slow economic growth, currency issues and heavy capital investment were noted as reasons for the results, which were a reversal after about eight years of strong revenue growth.

Smart was the best performer with revenue growth of 36 percent and  60.5 percent earnings growth. Though voice  revenue grew 19 percent, data grew 135 percent. Performance also was strong at the Robi, Dialog and Idea units, though 44 percent of revenue is earned at Celcom and 35 percent at XL.


Axiata's Malaysian mobile subsidiary Celcom fell 12 percent and revenue declined 3.5 percent.  

In Indonesia, XL Axiata increased its revenue by 10 percent to 23.6 trillion rupiah ($1.8 billion).

Axiata also recorded a strong performance at its Sri Lanka, Bangladesh and Cambodian subsidiaries, and solid contributions from its affiliates in India (Idea Cellular) and Singapore (M1).

Wednesday, February 25, 2015

Cablevision Makes "Big Shift:" Now is a Connectivity Company, not a "TV" Company

Cablevision Systems Corp. now says its most important product is Internet connectivity. The corollary is that, at some point, “I would suggest that we'll probably be agnostic about the video and not about the connectivity,” said James Dolan, Cablevision Systems Corp. CEO.

In other words, Cablevision has become more agnostic about linear video and over the top streamed video. Cablevision also is clear that high speed access is the foundation of the coming business.

“Connectivity has surpassed video as the primary product for a company like ours,” said James Dolan, Cablevision Systems Corp. CEO. “The consumer values the connectivity product more than they do the video product at this point.”

That is a “big shift,” said Dolan. If given a choice of linear video or Internet access, consumers “almost overwhelmingly will take the data product.”

And though it remains to be seen whether the current vision is the ultimate vision, Cablevision does not currently seen its Freewheel Wi-Fi- based phone service as a competitor to mobile service.

“We're not chasing the cellular markets to be a mobile phone provider,” said Dolan. “That's just not what the strategy for this product is about.”

And we should assume other products based on Wi-Fi access--inside and outside the home--are coming.

“There are products and services that will go along with that strategic position,” said Dolan. “The first of which you've just seen, which is Freewheel, but there will be others.”

“Wi-Fi is a critical part of our product strategy and one of our most important strategic assets,” said Dolan.

There are several important insights here. First, Cablevision sees itself as a connectivity company, not a video services provider.

Also, Wi-Fi is seen as a key building block for new services, beginning with a new Wi-Fi-based phone service, but likely to be followed with a mobile video consumption service.

Finally, Cablevision finally has launched a service it first envisioned in the late 1980s, namely an untethered communications service distinct from what we now know as “mobile phone service.”

An Illustration of Scale Benefits in Mobile

   U.S. Mobile Capex, Compared to Revenue source: TMF Associates
​Historically, North American mobile operators have invested twice as much capital per subscription as their global counterparts, according to analysts at ABI Research.

For 2015, ABI Research forecasts North America capital investment per subscription to be US$63. In comparison, Western European operators will invest US$34.

On the other hand, an argument can be made that U.S. mobile capex actually has been declining, on a per-account basis, for quite some time.

“In absolute terms, network investment has fallen quite sharply since 2004, and only rebounded partially in 2010 and 2011,” according to TMF Associates. “As a percentage of revenues, capex has fallen even further, and has roughly halved in the last decade.”

North American mobile operators invest 39 percent of capital on radio access networks. About 26 percent is targeted at in-building access.

ABI Research predicts North American and global mobile capital investment per subscription will decline, however, after a peak at US$224 billion in 2017.

The perhaps-key implication is that capital investment, as a percentage of revenue, is dropping, likely representing better efficiency because of scale.

      



Big Data is at Heart of Internet of Things

It is likely that big data, one of the big “hype” concepts just a few years ago, will find practical application as an integral part of developing Internet of Things and machine-to-machine applications that essentially gather data and then analyze it for application in real time, or near real time, to control the operation of devices, people and systems.

“IoT isn’t just about gathering data; it’s about using it to make better decisions — that’s the value of IoT, Verizon argues. “Regardless of whether the output is manual or highly automated, analysis of the data must be integrated into business processes. If the data is not actionable, by you or a third party, it’s not IoT.”

India Eliminates Fixed LIne Termination Rates, Cuts Mobile Termination 30%

Indian regulator TRAI has eliminated fixed line termination rates and cut mobile termination charges by around 30 percent. Notably, the reason for the fixed line termination charges for landline calls is that the fixed line voice business is in decline.

TRAI hopes the end of the 0.20 rupee ($0.003) termination rate for both landline-to-landline and landline-to-mobile calls will stimulate usage and subscriptions, and also lead to more investment in fixed networks.

TRAI also expects to reduce mobile network interconnection charges from 0.20 rupees to 0.14 rupees. That move is intended to stimulate mobile calling.

Those changes in the Indian market are not unusual. U.S. fixed network voice lines have been dropping since 2000. Globally, fixed voice lines have been dropping since about 2006.

Tuesday, February 24, 2015

Comcast's Next High-Growth Rate Business Will be Mobile

One way of illustrating the potential value of Comcast’s entry into the mobile business can be gleaned from looking at current revenue contributors, with their growth rates.

Of total 2014 Comcast revenue of $68.8 billion, $44.1 billion, or 64 percent, was generated by the cable communications business. Operating cash flow contribution from the cable communications segment was about six percent.

About 37 percent of revenue was generated by the NBCUniversal segment, which grew about 7.5 percent, overall. Operating cash flow from NBCUniversal was about 18 percent.

The main point is that gross revenue and operating cash flow from cable communications is tough, from a growth standpoint.

The consumer part of the cable communications business (triple play services) is growing primarily because of high speed access.

Video revenue for 2014 was up about one percent. Voice revenue was static at about 0.4 percent growth. High speed access was where the gains primarily were made, with growth of 9.5 percent.

Business services contribute about nine percent of cable communications segment volume, at about $4 billion in 2014. But business services grew at a 22 percent rate in 2014. In fact, it might be correct to say the newest product segments in the cable communications segment (business services in general, and mid-market services in particular) have the highest growth rates.

Comcast estimates it has reached about 25 percent penetration of the small business addressable market, but only about five percent of the addressable mid-market opportunity, according to Neil Smit Comcast Cable president and CEO.

So mid-market revenue “is growing at a increasing rate relative to SMB,” said Smit.

In 2013, Comcast “had literally zero penetration in the mid sized sector,” said Michael Angelakis, Comcast vice chairman and CFO.

The point is that Comcast has seen the fastest growth rates for successive new products.

That suggests the possible upside from Comcast adding a whole new product line in mobile services, especially if Comcast can position the service first as a way for Comcast customers to view their content, but then secondarily as a way to use voice, messaging and mobile Internet access.

Commenting on capital investments Comcast has made in Wi-Fi infrastructure, the emphasis has been on support for the video business.

“The real goal has been that our customers can access their video any time anywhere whether in the home or outside the home,” said Michael Angelakis, Comcast vice chairman and CFO. “If Wi-Fi can also develop into a different type of service then that’s an added benefit to the Wi-Fi investment.”

But Comcast’s Wi-Fi network now includes 8.3 million hotspots, said Brian Roberts, Comcast Chairman and CEO. “We think we are working on how we monetize that asset and bring it to market.”

“As you know we have MVNO relationships with Sprint and Verizon,” Roberts said, hinting at ways mobile service could be offered, but with much traffic offloaded to the Comcast fixed Wi-Fi network.

Adding mobile service would not only allow Comcast to offer a full quadruple play, but also would add a brand new product line with proven customer demand, but allow use of the new Wi-Fi hotspot network to lower costs.

80% of Smartphone Data Consumption Globally Now Uses Wi-Fi

Globally, Wi-Fi accounted for 80 percent of mobile and tablet data consumption, compared to data consumed on the mobile networks, at 20 percent, according to a new Mobidia report.

That explains the wide gap between reported “mobile data consumption” and actual end user data consumption on their smartphones.

Globally, smartphone and tablet users used in excess of 10 GB of data in December 2014, according to Ovum, up from about seven gigabytes in January 2014. That represents a 51 percent growth rate.

Apple iOS tablet users consumed about 12 GB, while Android tablet users consumed about nine gigabytes in December 2014, Ovum says.

Apple  iOS smartphone users consumed an average of about 11 GB of data in December 2014, just ahead of Android smartphone users with 10 GB. Consumers on Long Term Evolutin networks consumed even more data.

By December 2014, 4G Android smartphone users consumed 13 GB each month, dramatically higher than the 5GB/user/month of 3G Android smartphone users that month.

Wi-Fi has cemented its position as the dominant wireless access technology, with cellular playing a vital yet supporting role, Mobidia says, based on the results of a study conducted for Mobidia by Ovum.


“Wireless Internet access” is much more than “mobile access,” it is fair enough to note. Fixed wireless, satellite access and untethered access (Wi-Fi) are widely-used forms of Internet access beyond that used by mobile devices, connected to the mobile network.

It might also be fair to say that untethered access--which has been seen as a competitor to mobile access--might actually be emerging in precisely that way. “Cable operators, Internet giants, Wi-Fi-first startups and every cafe with a wireless router are all providers of wireless service,” Devicescape argues.

So a “wave of disruption” is coming, Devicescape argues.

Though it is reasonable to point out that Devicescape builds a business on the strength of Wi-Fi, especially the ability to create unified services out of a patchwork quilt of independent Wi-Fi hotspots, the notion that mobile service providers “must” or “should” incorporate Wi-Fi access into the overall fabric of connectivity choices is reasonable enough.

That, in fact, is assumed to be a key feature of future fifth generation mobile networks, and for simple reasons. Generally speaking, mobile network connectivity is best outdoors, less effective indoors, while Wi-Fi arguably operates best indoors, least effectively outdoors.

So mobile service providers must become “connectivity providers” using any available network resource, not “mobile access” providers, one might argue.

In part because as much as 70 percent of smartphone data is accessed using a Wi-Fi connection, mobile service providers must embrace connectivity by any available means, Devicescape argues.  

Devicescape calls that shift a move from “mobile” to “connectivity,” as users do not so much care about whether their access is provided by the mobile or the Wi-Fi networks. They only want to remain connected, at the best price, one might well argue.

That noted, Devicescape argues that 29 percent of mobile users never connect to their home Wi-Fi, while 53 percent keep Wi-Fi turned off when out and about. As a result, as much as 91 percent of public hotspot locations go unused.

Mobile Drives 72% of Webpage Views in India

Mobile continues to drive Internet access subscriptions globally, accounting for 72 percent of all India webpage views in 2014, for example. 

Globally, mobile devices account for about 80 percent of all traffic, such as sessions, many would argue. 

But overall data volume still is driven by fixed devices. How much longer that might be the case is an issue. Experts still disagree about how to measure the relative weight of mobile and fixed access, in part because usage varies so much by region and country

Slide046


Was 2014 an Aberration, or the New Normal?

Was 2014 an aberration in the U.S. mobile business, or a harbinger? The answer matters.

From 2010 to 2013, U.S. mobile data pricing (per unit sold) declined by only single digits year over year. In the first nine months of 2014, data pricing dropped by 77 percent, according to industry analyst Chetan Sharma.

The other key inputs to the business model shows escalating numbers as well. Average (mean) mobile data consumption increased to about 2 Gb a month. Sharma notes it took 20 years for consumption to reach 1 Gb per month usage levels.

The increase to 2 Gb took about a year.

In addition to plunging prices (less revenue per unit sold) and higher usage (more network cost), marketing costs have grown as competition has become more intense.

Overall U.S. operating expense rose 20 percent, year over year. Income was flat while earnings grew three percent.

Among the positives, total revenue grew 21 percent to almost $400 billion.

Among the negatives, voice revenue declined 15 percent, messaging by 16 percent and tablet subscriptions by four percent.

So was 2014 an aberration? In some ways, it has to be. Can mobile data pricing (per unit sold) continue to drop at breath-taking rates? Some might note the other key figure is the amount of incremental new buying that happens as per-unit rates drop.

Sharma argues that Internet access will not grow fast enough to offset voice revenue losses. Neither will device sales revenue or wholesale revenues grow fast enough, even in conjunction with mobile data revenue growth, grow enough to offset sharp declines in voice. That, in a nutshell, is the strategic challenge mobile operators face.  


Monday, February 23, 2015

Google Wallet to be Pre-Installed on AT&T, T-Mobile US, Verizon Android Devices

Google now is working with Softcard, the mobile payments service owned by AT&T Mobility, T-Mobile US and Verizon Wireless. As part of a new agreement between the mobile carriers and Google, the Google Wallet app, including the tap and pay functionality, will come pre-installed on Android phones (running KitKat or higher) sold by these carriers in the U.S. market  later in 2015.

As part of the deal, Google also is acquiring “some technology and intellectual property” from Softcard.

Google does not appear to have bought Softcard, nor do the mobile carriers appear to have sold the business to Google.

On the other hand, since October 2014, when Apple launched Apple Pay, both Google Wallet and Softcard, which have failed to get much traction, seem to have concluded they need to move faster to counter Apple Pay.

Cable TV Firms Will Not be Able to Replace Lost Linear Video Revenues with High Speed Access Gains

For the past decade, U.S. telco and cable TV revenue sources have been a game of musical chairs: give up the chair you have and hope you can grab another when the music stops.

And everybody is playing. Telcos have relied on mobile services to replace lost fixed network voice revenues.

Cable TV companies have grown by adding voice, high speed access and business segment customers.

Mobile companies have grown by adding tablet accounts on top of phone accounts, while adding Internet access revenues to displace dwindling voice and text messaging revenues.

But it can be a difficult game if the new revenue sources are smaller than the lost revenue sources, or if profit margins are significantly less robust. And that might be a growing issue.

“Gross margin dollars will be lost in video, and will not be replaced by faster growth in broadband,” said Craig Moffett, Moffett Nathanson analyst.

That might be why firms such as Comcast and Cablevision Systems Corp. are planning--or have launched--mobile services. The best way to add incremental revenue is to add one more new service to the bundle.

Comcast and Cablevision are using, or are expected to use, their own public Wi-Fi hotspots as the foundation of their mobile services.

Saturday, February 21, 2015

U.S. Mobile Market Structure Will be Shaped by Dish Decisions

As competitive as the U.S. mobile market has become--despite frequent protests that the market essentially is a duopoly--it almost certainly is going to become more competitive.

Strategically, Comcast is expected to enter the market, at some point. Cablevision Systems Corp. already has done so, though as a niche provider with greatest potential impact in its limtied home operating area. Comcast almost certainly will have to enter as a national provider.

Dish Network either must enter the market as an operator, or forfeit the rights to spectrum that presently accounts for as much as 80 percent of its total market value.

And then there are the other contenders, including Google, which it is believed soon will be entering the mobile market, and Apple, a perennial potential actor in the market as well.

But the biggest current question mark is what Dish Network will do.

Since nobody believes Dish will allow scores of billions in equity value to evaporate, Dish’s Long Term Evolution spectrum has to be put into service. That might consist of partnering with an existing carrier (normally assumed to be Sprint, T-Mobile US or Verizon) to create a wholesale-only operation, or a branded retail offering, or selling the whole spectrum portfolio or the whole company.

So the shape of competition--and the structure of the U.S. mobile market--rests on which alternative Dish ultimately pursues. Both Comcast and Google are expected to enter the market as retail providers, though each might pursue something of a niche approach.

If Dish decides to partner with one of the existing leading national operators, and become a wholesale provider, it will enable third parties, but not directly increase or decrease the number of national providers.

Should Dish decide to become a full branded retail provider, it might take a niche approach, heavy on video entertainment. But the manner of entry matters.

Some speculate Dish will try and buy T-Mobile US. In that case the U.S. market structure is not changed. Others believe Dish will partner with a network owner, but become a retail service provider, in which case the number of potential leading contenders in the market increases from four to five.

But Dish might also simply sell itself, in whole or in part, monetizing its spectrum holdings, but leaving the market structure unaffected. Some will argue the odds of a buyout that large, by either AT&T or Verizon, is extremely unlikely, given the debt loads both leading carriers now carry.

At some later point, such a purchase by Verizon--not AT&T--is possible, but not at the present time.

AT&T presumably already would have acquired DirecTV, which makes additional video assets superfluous. Nor does AT&T appear especially interested in investing much more capital in the U.S. market, given its Mexico market expansion or other potential growth moves outside the U.S. market.

Verizon must wait to get its debt burden reduced significantly before it can afford to consider an acquisition of all of Dish. A more limited buy of some significant spectrum assets is more feasible, but Sprint might be a seller of spectrum assets as well.

Complicating the analysis is the nature of potential Sprint and Dish Network spectrum assets. Some of Dish Network’s spectrum is in the 700-MHz band, while much is in higher bands near 2 GHz.

Some of the potential partners--especially Sprint and T-Mobile US--have plenty of spectrum in the 2-GHz range, so either have little need for additional spectrum in those bands themselves, but might be happy to earn revenue supporting Dish Network network operations.

Verizon, historically not a big player in the mobile wholesale market, does have some potential obligations to support wholesale operations by Comcast and other cable operators from which it purchased AWS spectrum.

Should that happen, Verizon might want the additional Dish spectrum to support its wholesale partners, in addition to its own retail requirements, in the future.

Complicating matters further is growing ability to use unlicensed spectrum to support and augment LTE, plus additional communications spectrum that will be added, at some point, on a shared basis.

More competition is coming. It just is hard to predict precisely what form that competition will take.

Study Finds Most Don't Understand Net Neutrality

A survey of U.S. consumers has found what you might expect, namely that few U.S. residents understand network neutrality. The Hart Research Associates survey found only 25 percent of respondents claimed they knew what network neutrality is.

Perhaps not surprisingly, 73 percent said they wanted disclosure of what the rules actually are, when told that “just five members of an unelected Federal Communications Commission will decide the future of the Internet without providing an opportunity for the public to see and understand the regulations prior to a vote.”

The wording of the explanatory note, most might suggest, was not “neutral.” So it might not be surprising that 80 percent of respondents wanted full disclosure of the ruling before a vote is taken.

Just 32 percent of respondents thought regulating Internet access like telephone services would be helpful.

Business, or Consumer, Few Actually Benefit from Really Fast High Speed Access

With the caveat that a business customer’s use of bandwidth differs from the pattern typical of a consumer customer, small business customers of Cogent Communications tend to use about 12 Mbps of the 100 Mbps services bought to replace T1 connections of about 3 Mbps, said BTIG researchers.

According to Cogent, only about 12 to 24, out of perhaps 17,400 customers ever have reached 50 percent utilization of the 100 MB pipe.  Likewise, customers who buy gigabit connections have usage that does not likely differ materially from 100 MB customers, according to Cogent.  

One might well argue that consumer consumption is growing faster than business customer usage, certainly. But Sandvine data suggests U.S. median household data consumption over a fixed network connection is about 20 Gb a month. Granted “Gbps” is a measure of speed, while “Gb” is a measure of consumption, but monthly consumption of 20 Gb does not suggest most households likely are taxing their access downlinks.

To be sure, households with faster connections tend to consume more data. But that might be because households consuming more data disproportionately buy the faster connections. As more locations are able to use connections operating from 40 Mbps up to 1 Gbps, we should get a better idea of how much a “typical” user consumes, when access speeds exceed the ability of far-end servers to respond.

A study by Ofcom, the U.K. communications regulator, suggests that beyond about 10 Mbps, local access speed is not the experience bottleneck.

The study found that “access speed” matters substantially at downstream speeds of 5 Mbps and lower. In other words, “speed matters” for user experience when overall access speed is low.

For downstream speeds of 5 Mbps to 10 Mbps, the downstream speed matters somewhat.

But at 10 Mbps or faster speeds, the actual downstream speed has negligible to no impact on
end user experience.

Since the average downstream speed in the United Kingdom now is about 23 Mbps, higher speeds--whatever the perceived marketing advantages--have scant impact on end user application experience. Some 85 percent of U.K. fixed network Internet access customers have service at 10 Mbps or faster.

Investing too much in high speed access is, as a business issue, as bad as investing too little, one might argue.

Average access speeds in the United States are 10 Mbps, according to Akamai. Average speeds are 32 Mbps, according to Ookla. Another study shows that average Internet access speeds in the United Kingdom and United States are equivalent, in fact.  

The point is that, in terms of user experience, faster marketed speeds (gigabit, 100 Mbps) actually do not improve end user experience.

As someone who recently was able to upgrade from about 15 Mbps to 105 Mbps, I would confirm, as an end user, that the upgrade has made no apparent difference in my browsing experience.

For that reason, I will not be buying a gigabit access connection, which I could do. There being no apparent change in experience at 100 Mbps, I cannot see the advantage of upgrading further, to 1 Gbps.

"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...