Tuesday, September 23, 2014

Mexico to Consider Wholesale-Only LTE Model

In markets where there is separation between physical networks and retail sales of communications services, the “dumb pipe” issue is germane.

Where there is structural separation of the network function and all retail and branding operations, what is access but an essentially commodity-like pipe function, equally available to all contestants?

In such scenarios, the value of the product is determined not by the features of the wholesale network, since every contestant gets the use of that resource, at the same wholesale price.

In that case, differentiation on wholesale-provided features is not possible, and pricing differentiation will be limited.

Uniqueness therefore has to be added using elements not sourced from the wholesale network infrastructure provider.

That is not to say the network-based services are actually “dumb,” only that network-derived features will be tough to differentiate. But that’s essentially the business problem illustrated by the phrase “dumb pipe.”

In a number of world markets, service providers are going to have to learn how to manage that issue, as a number of countries are taking a “wholesale-only” approach. Some, as in Australia, are doing so to support fixed network services.

Others are doing so to expedite availability of Long Term Evolution fourth generation networks. As a result, new thinking about sources of value is likely to happen, since retail service providers will not be able to differentiate on the extent of coverage or features of the mobile network.

There might be some ability to differentiate on price or packaging, but that will be limited. Other features and values not related to the physical network and its features will be necessary.

In a way, that makes wholesale-driven markets more susceptible to “dumb pipe” commoditization, even if functional structural separation also means retail contestants avoid the capital investment that otherwise would have been required.

In a way, that means less concern about raising capital, but much more thinking about how to create uniqueness and value. And that might happen in a growing number of markets.

In a number of cases, entire new networks, owned by the government, either are proposed or underway. Australia’s National Broadband Network provides an example in the fixed network segment of the business.

But a group of potential investors also now has submitted a bid to the government of Mexico  to finance the construction of a new wholesale-only mobile network.

That is similar to the situation in Rwanda, where 4G spectrum was donated--not auctioned--to a an entity charged with building a national LTE network. In Rwanda, KT Corp. was selected by the Rwandan government to build a national LTE network, known as olleh Rwanda Networks, (oRn).

The new infrastructure company oRn will operate exclusively in a wholesale capacity, providing services to retail service providers. And it appears that as many as nine other African nations are considering doing something similar.

Nigeria also appears to have taken the wholesale-only LTE approach.

Kenya also is interested in a wholesale-only LTE network approach. Existing mobile service providers, to nobody’s surprise, have not been entirely sure they want to operate under such a structure.

The larger mobile service providers, including Safaricom, would much prefer to own their own spectrum. So it still is not clear whether the wholesale approach will win.

Safaricom serves over 66 percent of mobile users in Kenya, so its participation in any wholesale LTE plan is likely crucial. And Safaricom might well continue to hold out for getting its own spectrum.

Safaricom’s reluctance again focuses attention on the strategic value of spectrum assets, at least where multiple licenses are granted.  

In markets where a single wholesale entity controls all LTE 4G spectrum, the access network itself arguably does not confer competitive advantage.

It’s another example of how “dumb pipe” remains a crucial issue for telecom service providers.

In Mexico, it is unclear whether the new proposal will succeed. The investing group still has not secured all the capital required. The government hasn’t yet agreed to donate spectrum.

But if Mexican regulators agree, the new wholesale company will have exclusive use of 90 MHz of spectrum freed up in the transition from analog to digital broadcasting.

"Without a shared network ... it will be impossible to have sufficient telecoms service coverage, keeping our country behind," Communications and Transport Secretary Gerardo Ruiz Esparza said earlier this year.

That Mexico’s mobile market is in need of competition is hard to contest. Mexicans pay the highest prices for the slowest median broadband speeds in the 34-nation Organisation for Economic Co-operation and Development and mobile penetration is among the lowest in Latin America.

Up to this point, America Movil has had 69 percent share of the mobile market. Number-two provider Movistar, owned by Telefonica, has had 19 percent market share.

Whether Mexico will go the wholesale-only route remains to be seen.

Monday, September 22, 2014

Is AT&T DirecTV Acqusition a Big Mistake, a Clear Win, or a Deal of Unknown Ultimate Impact?

Some critics believe AT&T should not attempt to buy DirecTV. Some might say the deal was a reaction to the Comcast proposed acquisition of Time Warner Cable and will not provide enough synergies to drive value. In that view, AT&T should spend its acquisition or network capital elsewhere.

Some question the amount of potential acquisition savings AT&T has claimed. But that might not even be significant. The issue is that AT&T might wind up in a situation where it has to pay out 70 percent to 80 percent of its cash flow to cover its dividends.

For that reason, some think AT&T simply shouldn’t do the deal. So far, there is no evidence AT&T will withdraw. In fact, AT&T already seems to have reached agreement with regulators about what AT&T has to do to gain antitrust clearance.

Those who oppose the deal are left with the hope regulators will block the deal, but that seems unlikely, in view of the news that antitrust concerns have been dealt with.

But what if “synergies” are not the issue, or even video scale? What if AT&T actually believes DirecTV offers a high cash flow opportunity, even if it might be a slowly declining business, and that the cash flow is reason enough to do the deal?

AT&T would not be the first company to find it must finance both high investments in its core business, as well as pay significant dividends.

In that view, whatever AT&T might do, access to high cash flow from DirecTV helps by supplying cash for dividend payments, while the company continues to invest in its gigabit networks and other new lines of business.

To be sure, some will make the argument that AT&T essentially is not telling the truth; that it will eventually do something other than what it now says it will do.

An independent DirecTV might actually already believe it has passed the highpoint of its subscriber adoption, if it remains an independent company.

But AT&T might argue it has a solution for that problem, namely the ability to bundle DirecTV with mobile service, fixed Internet access and voice. In that scenario, might DirecTV’s eventual decline is much more gradual?

It’s a reasonable theory, if still a theory. Keep in mind that virtually everybody believes linear video, as an industry, already has passed its peak. What forecaster, inside the companies or outside them, really believes the market can grow from here?

Rather, the strategic challenge is to prolong product demand as long as possible, as profitably as possible.

AT&T and DirecTV might believe bundling opportunities will play a big role in that regard.

Some argue there are synergies.  Certainly AT&T has suggested it will benefit to some extent by becoming a more-sizable buyer of content for its linear video services, on the strength of DirecTV’s mass.

But much will hinge on what conditions the Federal Communications Commission or Department of Justice might attach--and to which AT&T already has agreed--in exchange for approving the merger.

If past precedents provide guidance, that will mean a period of perhaps three years when existing packages and prices for current customers are protected from change.

In a drastic scenario AT&T obviously does not expect, AT&T might be required to divest video subscribers in any areas where its U-verse services are offered. That seems an unlikely FCC or DoJ objection, but it is possible.

Still, it seems unlikely AT&T would be willing to buy DirecTV for its stated price if AT&T believed the value of DirecTV would be reduced by as much as 5.7 million video accounts that would have to be divested.

About as thorny an issue as that is which network would support the divested customers. AT&T, it seems likely, would not want to support a new third-party owner of divested U-verse video accounts, running over AT&T plant. Nor would AT&T seemingly gain so much contract scale were it to have to divest its 5.7 million U-verse video accounts.

It therefore is hard to believe significant divestitures are contemplated.

Assuming few if any accounts actually must be divested, there is the matter of potential synergies. AT&T might, some think, simply continue to operate U-verse video on the fixed network and DirecTV as separate businesses.

There are no network synergies there. And some might argue there could be additional costs, particularly if AT&T has to do something major with the DirecTV or U-verse decoders to harmonize features and user experience.

DirecTV decoders might ultimately be outfitted with Long Term Evolution capability for return path signaling, for example. Depending on the number of box swaps, that could represent a rather large capital outlay, at $400 per customer, on average. Or pick some other number, if you like.
The point is that synergies might be few. And AT&T might still want to do the deal.

Any bundled offers would necessarily be a billing and marketing issue, with actual service provided by as many as three separate networks--satellite, fixed and mobile networks.

It might not be as elegant as a single fiber-to-home platform, but it can work. Service providers already do such things.

Still, there are other currently-imponderable longer term decisions. Perhaps half of AT&T high speed access customers bundle video with broadband access.

In the first quarter of 2014 AT&T had 11.3 million U-verse broadband customers and 5.7 million U-verse TV subscribers.

Compare that to Verizon Communications, which sells a FiOS video unit to about 80 percent of customers who buy FiOS broadband.

Verizon had 6.17 million FiOS broadband customers as of March 31, 2014, plus 5.32 million pay-TV subscribers.

If regulators approve the DirecTV purchase, AT&T says that it will still offer U-verse-branded TV to consumers in areas where its network has been upgraded to enable the service.

What happens in the future is probably the bigger issue. "Longer term, I think they will look to sell a bundle of satellite and U-verse data: it's more efficient and a better service," argues  UBS analyst John Hodulik.

Whether that is “better” is a judgment call. But that approach could have implications. Some might argue AT&T will simply slow down its bandwidth upgrades. If it does not have to support video services over its fixed network, virtually all the physical bandwidth could be devoted to Internet access.

Cable companies face the same issue: they have to apportion bandwidth to support both video and Internet access. One of the advantages of switching to all-digital video delivery is that it is more efficient, in terms of bandwidth consumption, and essentially allows cable operators to use more of the network to support high speed access.

At some point, more than three years out, one could conceive of a move by AT&T to stop delivering video over its U-verse network. That would be disruptive, and could lead customers to churn off U-verse in substantial numbers, so one would think that is an unlikely plan.

U-verse video, assuming an average $80 per unit, represents $960 a year of gross revenue. At At 5.7 million units, that is $5.47 billion a year in annual revenue.

Of course, one can imagine other scenarios. AT&T could try and induce customers who buy U-verse video to buy DirecTV instead, reducing its exposure to the loss of U-verse customers, were it ever to decide U-verse video was not warranted.

That migration strategy would reduce exposure, if AT&T did decide to abandon U-verse video at some point. A couple of scenarios could drive that decision.

Linear video subscriptions could take a suddenly sharper downward path industry wide, shrinking the total linear video market.

Or AT&T might someday conclude that the profit margin on high speed access is so much higher than linear video that the business case for devoting nearly all bandwidth to high speed access becomes more compelling.

It is less crazy than it sounds. Cable operators already can model a future business where demand for linear video is much lower, and demand for high speed access is much higher.

Verizon Communications executives say in public that the profit margin on FiOS video really isn’t that great.

There are probably more scenarios than we can even imagine right now.

Thursday, September 18, 2014

Are Access Networks Still So Valuable?

A deregulation of fixed network voice prices in some European Union countries, a possible decision by Verizon to sell its tower networks and objections to AT&T’s purchase of DirecTV all have one common thread, namely an estimation of the value of access networks, to whom.

Where national regulators agree, fixed network service providers will be able to set prices for retail and wholesale voice services dictated only by an estimation of demand and value, and will not be price capped.

Of the customer segments, it arguably will be more significant that wholesale rates can be set at market rates, since in many European markets, competitors buying wholesale services from incumbents represent half or more of all accounts on the incumbent networks.

The potential shift to higher rates, particularly the wholesale rates, could essentially raise the value of the incumbent access networks, since wholesale revenue will rise.

In the U.S, market, a lifting of mandatory pricing rules essentially doomed most competitive voice providers relying on mandatory wholesale access at significant price discounts (of as much as 40 percent).

At least partially as a result, facilities-based cable TV operators quickly became the dominant class of competitors to telcos in the fixed network voice services business.

But aren’t higher wholesale or retail prices bad for consumers in the near term? It depends. It is conceivable that some marginally profitable suppliers will find their business models squeezed, and retreat from the market.

But facilities based cable TV operators will have so such problems. What could happen is that market share formerly held by wholesale-based providers is taken by facilities-based cable operators.

Regarding the potential sale of Verizon cell sites, possibly 12,000 to 15,000 discrete sites, one might initially argue such a move would suggest ownership of access facilities is not so important in the mobile business.

Actually, ownership of “facilities” remains of high strategic value in the U.S. mobile business. But the relevant facilities consists of rights to use spectrum exclusively, not the ownership of towers.

And access facilities nearly universally is deemed valuable, even in a business context where wholesale-based competition is possible.

No U.S. cable TV operator, for example, ever has proposed giving up its network and providing services over a leased network.

Most tier one fixed network telcos have resisted older common carrier rules that would grant widespread mandatory access to any new fiber access network they might build, as well. That is one way of suggesting such networks have important scarcity value.

Mobile service providers, one might say, have more nuanced views of the value of mobile tower assets, simply because the truly strategic asset is the exclusive right to use spectrum.

That is why in some markets, competitors share the cost of a single tower network, and might or might not share radio resources.

Sprint and AT&T have sold off their tower networks and lease access. Now Verizon might be willing to consider doing so as well. The point is that mobile executives do not see the same scarcity value in mobile networks as do fixed service providers.

A third angle is that some who object to AT&T doing any number of things often suggest that AT&T needs to keep sinking capital into copper networks to support voice.

That might seem odd, since many who criticize AT&T for not upgrading to optical fiber fast enough also say they want AT&T to spend more capital maintaining a copper network.

AT&T would prefer to make a faster transition to new that are either all fiber or fiber-reinforced in ways similar to cable hybrid fiber coax networks.

That likewise speaks to the perception of value. AT&T believes its vast fixed network consumer and business access networks have scarcity  value, but the copper networks arguably have less value than optical networks, in no small measure because wholesale obligations for optical access have fewer mandatory wholesale obligations.

Competitors, on the other hand, would prefer that AT&T be forced to maintain its copper facilities, since the competitors have wholesale access beneficial to their business models.

Paradoxically, in many quarters there is pressure on AT&T to keep investing in aging copper plant at the same time AT&T is asked to invest faster in the next generation optical and optical-reinforced networks. But one goal (keep investing in copper plant) drains capital from achieving the other goal (invest faster in optical-backed facilities).

So though it might occasionally seem as though access networks do not represent high business value because such facilities are scarce, the reverse is true.

Access networks still confer high business value because they remain truly scarce assets.

Connected Car is an Early Candidate for Mobile-Supported IoT or M2M Because Revenue Model is Clear

Sometimes the actual details of a product growth forecast are less important than the depiction of direction. That probably is true for any forecast of revenue potential for sensor-based applications of the "machine to mahcine" or "Internet of Things" sort.

Nor, at this point, do our efforts to define the difference between M2M and IoT so important. The direction is the main thing. The big take-away from a forecast produced by Telco 2.0 Research is the potential growth of connected sensors or other monitoring devices.

Connected things potentially will rapidly outstrip phones, PCs or tablets as devices using moible and other networks by about 2020, representing more than half of all connected devices by 2020. 

Virtually all forecasts show exponential Internet of Things growth from now to 2020, outstripping additions of mobile phone, connected PC or connected tablet accounts. 

What remains unclear is how revenue oppotunities might evolve, and which networks and service providers stand to benefit most. If most of the devices or sensors use Wi-Fi, incremental access revenue will be small to moderate.

If most of the devices or sensors rely on mobile connections, mobile service provider service revenues would get a big boost, albeit with per-sensor revenues far smaller than generated from phone connections. 

But it is hard to ignore the potential volume of new connected devices. What remains unclear the magnitude of the growth, as current IoT or M2M forecasts vary by an order of magnitude. 

The other issue is the fragmented nature of the "market," which might be said to consist of scores of potential markets, ranging from automotive to logisitics to health care to smart building apps, as well as security or logistics, for example. 

So far, utility apps have been in the forefront, in part because of government mandates. Right now, "connected car" is getting lots of attention. Among the reasons is the certainty that the connections will have to be provided by a mobile network of some kind, making the revenue opportunity clear for a mobile service provider. 

In other stationary apps, it isn't so clear that mobile networks are the only option, or the best option. 



source: Telco 2.0 Research 

Wednesday, September 17, 2014

"Skinny Bundles" Illustrate Growing Revenue Instability in Linear Video Entertainment

Smaller bundles of channels increasingly are seen as one way for linear video providers to provide consumers with more-affordable bundles of channels, while giving distributors more leverage in negotiations with program suppliers.

In the U.S. market, service providers are looking at smaller bundles of perhaps 20 channels, offered as a “streaming only” package. Dish Network and Verizon are the foremost proponents of such offers, at the moment.

Whether “skinny basic” packages can be created for linear video packages is the issue. Current programming contracts make that difficult, as standard contracts typically require that channels be included in whatever tiers are most popular with consumers (measured by number of subscribers taking each bundle).

In Canada, the Canadian Radio-television and Telecommunications Commission is considering mandating offering of lean programming packages that cost less, possibly by stripping out U.S. channels from the low-cost bundles, or even making channels available on a full “a la carte” basis.

Such discussions have been active in Canada since at least 2012, when the CRTC issued rulings related to bundling rules.

Since then, the CRTC has a number of options for containing consumer costs, including full a la carte offers, a move with unsettling implications for service providers and customers.

Nobody knows precisely how pricing might change in an environment where consumers can buy any single channel, on a stand-alone basis. That would create huge instability for service providers, who would likely lose the ability to predict future revenues.

On the other hand, consumer impact might be highly disparate. Some consumers could save money, while others find they wind up paying the same. And some consumers might even pay more. Everything hinges on which channels consumers really want, how those channels can be sold, and what rules might be created pertaining to bundling.

The present proposed changes in CRTC rules illustrate the amount of instability that now seems to be growing in traditional service provider businesses. Nobody really knows how distributor revenues might change if a la carte rules are imposed, or what consumer reaction might be.

Observers watching for signs of major change also are paying attention to what Time Warner might decide to do with its HBO service.

Some day, it has been reasonable to predict, one content owner would break with precedent and offer it’s content direct to consumers, over the top, without the requirement to buy that same product as part of a linear, bundled video subscription.

Some have predicted HBO would be first to do so, as HBO already does so in the Nordic countries, and has sold its service as an over the top product since 2012.
That logically suggests HBO would be willing to do so elsewhere where such a tack might make HBO’s owner Time Warner, more money than restricting over the top sales.

Time Warner has gradually warmed to the idea of doing so even in big legacy markets such as the United States, juding by public statements over the past few years.

As recently as 2013, though, top HBO executives continued to insist there was no business model in the U.S. market.

But business viability is coming closer, Time Warner CEO Jeff Bewkes indicates.

“Up until now, it looked to us as though the best and main opportunity was to focus on improving the penetration, the offering, the servicing, the interface, the monetization of HBO through the existing affiliate system,” Bewkes said.

“So now the broadband opportunity (over the top delivery) is getting quite a bit bigger, and the ability of the plant to deliver something robust is getting stronger,” said Bewkes.

That doesn’t mean the tipping point has been reached. It hasn’t.

But the new language does suggest the advent of over the top HBO streaming is closer than it has been in the past.

“We’re seriously considering what is the best way to deal with online distribution,” said Bewkes.

HBO is a logical candidate to move first for several reasons. HBO historically has been marketed as a premium add-on to the basic linear video subscription.

In other words, customers had to first buy a basic cable subscription of some sort before they were eligible to buy HBO or other “premium channels.”

In part for that reason, premium channel buy rates always have been a fraction of basic plan purchases, although it can be argued that aggregate purchases of premium channel subscriptions amounts to about 88 percent of basic tier buy rates.

In 2013, about 23 percent of U.S. households purchased Showtime, while about 29 percent of households bought HBO.

Purchases of  Cinemax penetration stood at nearly 14 percent while Starz penetration was 22 percent.

So the challenge for HBO remains: how much incremental purchasing could it gain by offering an over the top product, and how much might it lose in support and revenue from video distributors.

At least so far, the risk has been deemed too high. But thinking is evolving. The big issue is what happens once HBO really does make a change.

At the same time, Canada might be on the cusp of a rather significant change affecting the widely-viewed basic channels. Even if full a la carte does not become law, the way the skinny basic tiers are created could have significant impact on programmers.

If excluded from skinny basic packages, and if those skinny basic packages become popular, some programmers might find they cannot afford to remain in the market. That obviously would reduce programming diversity and end user choice.

Like streaming alternatives, a la carte and skinny basic illustrate the coming challenge for video suppliers and distributors. What are the odds fundamental changes (a la carte or streaming) will enhance revenues for programmers or distributors, rather than leading to lowering revenue?

To be sure, the existing ecosystem participants might prefer a revenue-neutral or revenue-enhancing solution. But it simply remains unclear whether that is a reasonable expectation, if or when unbundling is widespread.

Tuesday, September 16, 2014

U.S. Cable and Telco Businesses Will be Tougher, Soon. The Only Issue is How Much Tougher

If you have followed U.S. communications policy for any period of time, you know that regulatory policies blow hot and cold, for industries and segments within industries subject to regulation, from time to time.

Not often in recent decades have U.S. communications providers faced such strong negative headwinds as they do at the moment.

The U.S. Federal Communications Commission  says it might consider regulating high speed access as a common carrier service. Most observers argue that would immediately decrease investment by telcos or cable companies, as the expected financial return would drop, while operating costs might well rise.

Then there is the impact of new network neutrality rules on mobile and fixed Internet service providers, generally expected to reduce revenue opportunities for the affected firms.

Add also the growing demand for government-owned or run access networks, as well as FCC arguments that it has the authority to overrule any state-passed laws concerning government networks.

Connecticut, for example, appears to be among the U.S. states that wants to encourage more public entities within the state to pursue building of new gigabit access networks.  

On September 2, 2014, the City Council of North Kansas City approved a 10-year agreement with local company DataShack for the operation and maintenance of the city's fiberoptic network, known as liNKCity.

DataShack will operate and maintain the fiber network, but the city will continue to own the network.

Under the agreement, DataShack will collect revenue for broadband services sold to businesses.

But as of January 1, 2015, DataShack will provide existing and new residential broadband customers with gigabit service for an installation fee of $300 (or $100 for 100Mbit/s service, or $50 for 50Mbit/s service).

Residents in the city of about 4,000 people won't pay any more for service for the duration of the 10-year deal. In other words, gigabit access will be free for 10 years.

As part of the agreement, North Kansas City will share profits and losses equally with DataShack, with any potential losses for the city capped at $150,000, including capital investment.

DataShack will assume all costs associated with providing free gigabit services, reports indicate.

Some might welcome such a development, in the hope it will reduce access provider “power” within the communications ecosystem. That arguably could happen. But so could a massive “capital strike” by investors that would cripple service provider efforts to rapidly upgrade access infrastructure.

It’s a giant game of chicken with the U.S. Internet access infrastructure at stake.

Some will simply quip that if the telcos and cable companies refuse to invest, let them go bankrupt. Others will take up the challenge. Maybe. But fixed access networks remain hugely expensive undertakings with a growing amount of risk.

In the end, it is likely ISPs will operate in a more-constrained regulatory environment and a more-competitive business environment.

There won’t be much alternative but to take whatever measures are necessary to restructure business plans so incumbent providers remain competitive with attackers. You know the story well enough to predict what will happen to the incumbents.

Profit margins will fall. Revenue growth will slow. Operating costs will have to be reduced.

A tougher environment is coming. The only issue is how much tougher it will be.

Redefining "Broadband" is Only One of Several Big Challenges Service Providers Now Face

Among other potential changes in U.S. policy pertaining to high speed access is the question of what numerical value should be set to define what “broadband” actually is.

At the moment, the Federal Communications Commission uses a functional definition of a minimum of 4 Mbps in the downstream direction, and 1 Mbps upstream.

The definition matters because universal service support and other data collection efforts hinge on use of the definition.

FCC Chairman Tom Wheeler says 25 Mbps is “table stakes,” while universal service should be based on a definition of at least 10 Mbps for Universal Service Fund activities. In other words, the FCC chairman thinks classic “Ethernet” speeds of 10 Mbps should now be the minimum any Internet service provider must provide, in a rural area, to qualify for universal service support.

That has clear financial implications for at least some rural ISPs, and also creates a sort of “speed umbrella” for all competitors in rural areas. In other words, in many markets, ISPs will have to contend with a speed floor at 10 Mbps, or whatever speed the FCC might mandate as the minimum for getting universal service funds.

To be sure, floors are not ceilings. Some service providers might already be selling services at speeds of 25 Mbps or much higher.

The new definition might, or might not affect mobile service providers as well. In some cases, mobile “broadband” might suddenly become less widespread, if the definitions used for fixed and mobile networks are the same.

Even Long Term Evolution networks might not meet a 25 Mbps standard immediately. On the ohter hand, some providers, such as T-Mobile US, might welcome a chance to upstage its major competitors, as T-Mobile US already is operating LTE networks operating as fast as 100 Mbps.

The other important non-financial angle is that resetting minimum definitions from 4 Mbps up to 10 Mbps or 25 Mbps will likely have implications for the integrity of record keeping, creating a statistical anomaly in the year the new and different definition is used.

It is possible that the percentage of U.S. “broadband users” will decrease when the change is made.

Over the long term, such periodic revisions will make sense for government regulators. When large numbers of U.S. consumers are able to buy service ranging from 100 Mbps to 1 Gbps, keeping a 4-Mbps definition is silly.

On the other hand, some service providers--satellite and fixed wireless come to mind--will be quite challenged as overall speeds climb. For satellite providers, network architecture is the limit. For fixed wireless providers, available spectrum is the limitation.

In a market where 1-Gbps or 300 Mbps access is common, 15 Mbps services will seem like dial-up access.

Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...