Monthly Archives: August 2020

AT&T Argues for Broadband Reform

Ed Gillespie, the Senior Executive Vice President of External & Legislative Affairs at AT&T posted a policy position on the AT&T website that argues for major policy reform to better bring broadband to low-income homes and rural areas.

It’s hard for any broadband advocate to not agree with the suggestions Mr. Gillespie is making:

  • He wants Congress to finish funding the new FCC mapping program to identify homes without access to broadband.
  • He supports additional broadband grant funding for programs like the $20 billion RDOF grants.
  • He supports Lifeline reform and says that it should be as easy for low-income homes to apply a Lifeline discount as it is to use a card to buy food from the SNAP program.
  • He thinks funding should be increased for the Lifeline program and should be funded by Congress rather than funded through a 26% tax on interstate telephony.

I hope AT&T is serious about these proposals because having them lobby for these ideas would help to move the needle on digital inclusion. It’s just odd to see these positions from AT&T since they have spent a lot of effort and dollars arguing against some of these policies.

Mr. Gillespie complains that a lot of the current $9.25 Lifeline discount program is used by MVNOs and other carriers that have not built networks. That’s an ironic argument for AT&T to make since the company has done it’s best to walk away from the Lifeline program. AT&T no longer offers Lifeline in 14 states – AL, AR, FL, IN, KS, KY, LA, MS, NC, NV, SC, TN, and WI. AT&T still participates in Lifeline in 6 states, but only because those states refuse to allow the company to exit the Lifeline program.

Of course, this would not be an AT&T policy paper if the company didn’t pat itself on the back a bit. Mr. Gillespie brags that the ISP networks in the country weathered the big increase in web traffic due to the pandemic even though predictions were made that networks would collapse. He claims that AT&T made it through the pandemic due to light touch regulation. The fact is, once it was understood that the new traffic on the web was coming during the daytime when the network wasn’t busy, I don’t know any network engineer who thought that the web would collapse. I also wonder why AT&T would claim to have weathered the pandemic well – I would challenge AT&T to bring forth happy customers using AT&T DSL and ask for their testimonials on how the AT&T network enabled multiple people to work from home at the same time.

Mr. Gillespie is also calling for an expansion of the concepts used in the RDOF grants. Those grants provide funding for new broadband networks in rural areas that have the worse broadband. Before supporting an expansion of that grant program, I think many of us are holding judgment on the RDOF reverse auction process. While I think it’s likely that there will be beneficial grants given to those willing to build rural fiber networks, I also fear that a huge amount of these grants are going to be wasted on satellite broadband or other technologies that don’t bring rural broadband in line with urban broadband. I’m not ready to bless that grant program until we see how the reverse auction allocates money. I also can’t help being suspicious that AT&T’s position in favor of more grants reflects a hope to win billions of new grant dollars.

Interestingly, even though he never says it, the reforms that Mr. Gillespie is asking for require new broadband regulation. I’m sure that Mr. Gillespie must realize that bills needed from Congress for these reforms are not likely to stop with just AT&T’s wish list. We are long overdue for a new telecommunications act that brings broadband regulation in line with today’s reality. The last such law was passed at a time when people were flocking to AOL for dial-up access. It’s highly likely that new telecom legislation is going to go beyond what AT&T is calling for. It’s likely that new legislation will give some broadband regulating authority back to the FCC and will likely include some version of net neutrality. It’s ironic to see arguments for a stronger FCC when the FCC walked away from regulating broadband at the urging of AT&T and the other giant ISPs. Perhaps even AT&T knows it went too far with deregulation.

Being Stingy with Broadband Speeds

I’ve never understood ISPs that build fiber networks and then sell small-bandwidth products. The fiber technologies in place today can easily provide gigabit speeds to every customer without straining the network. The cost of providing 10-gigabit electronics keeps dropping and is now only a few hundred dollars extra per customer. Why would a fiber-based ISP have speed tiers that provide 50 Mbps?

I’ve found that it’s not unusual for an ISP with a low-bandwidth product on fiber to also charge a lot for gigabit bandwidth. There are a number of ISPs that charge $150 to $200 for a residential gigabit bandwidth product.

Bandwidth pricing philosophies differ around the industry. There are ISPs like Google Fiber, and Ting that only offer gigabit broadband. These ISPs are declaring that their fiber is a faster technology and they are marketing based upon that technology advantage.

The big ISPs in the country have trained the public that extra bandwidth is expensive. The cellular companies are the king at this game and will sell an extra gigabyte of usage for as much as $10. The big ISPs like Comcast and AT&T charge a lot for any customer that exceeds an arbitrary data cap. These pricing philosophies make a lot of money for the big ISPs, but this pricing conveys the false message that extra customer usage drives up costs for an ISP, and that customers that use more data ought to pay more.

Anybody who understands how ISPs operate realizes that there is little or no incremental cost for a given customer to use more bandwidth. It seems counterintuitive, but a household that uses a terabyte of download for a month doesn’t cost the ISP any more than the customer that uses 200 gigabytes per month – all due to the way that ISPs pay for wholesale broadband. ISPs buy enough bandwidth to satisfy the busiest hour of the day – and the rest of the day a lot of the bandwidth sits unused. There was zero incremental cost to ISPs for wholesale broadband when their customers started downloading and uploading a lot more data during the daytime due to COVID-19 – because the daytime usage still didn’t exceed the evening busy hour.

An argument can be made that faster speeds are more efficient for an ISP.  Consider the difference between two customers that each download a 1-gigabyte data file. A customer with a 50 Mbps product is using the network, and potentially interfering with other traffic for twenty times longer than the customer with a gigabit bandwidth product. Faster speeds reduce collisions between data streams, and a fiber network with fast customer speeds is significantly more efficient than one with slower speeds.

This is not to say that I’m advocating that ISPs should sell only the gigabit product since that is the most efficient use of a network. I think ISPs with only a gigabit product are leaving revenues on the table – unless the product is priced low enough to be affordable for everybody. I think companies that only a gigabit product at $70 or $80 are pricing out of the financial reach of many homes.

I get to peek behind the curtain of a lot of ISPs, and I know that an ISP with a smart tier of products can have more customers and more revenues than the ISP with only one product. I’m positive that an ISP with a $60, a $70, and an $80 product will do better than an ISP with only a $70 gigabit product.

The COVID-19 pandemic has finally forced the industry to confront broadband affordability. Even in markets where there is fast broadband available, we found out during the pandemic that there are a lot of homes without broadband for students because their parents can’t afford it. ISPs have not cared much about the homes that can’t afford broadband – and in most markets, that’s anywhere from 10% to 40% of the market. ISPs have been happy selling expensive broadband to those that can afford their prices and have given little thought to those that can’t.

I am truly puzzled why ISPs with fiber networks have broadband products between 25 Mbps and 75 Mbps. That’s like buying a race car and driving it on the freeway at 25 miles per hour. A cable company is not afraid of a competitor that wants to fight the market battle at speeds the cable company can match. The cable companies are afraid of ISPs of affordable symmetrical data products they can’t match.

A Fresh Look at Cord-Cutting

Roku undertook a survey that took a deep dive into cord-cutting and interviewed over 7,000 homes in March. The overall conclusion of the survey is that cord-cutting is accelerating in 2020. The survey was done at the beginning of the pandemic, and overall industry statistics for the second quarter make it sound like cord-cutting exploded in the second quarter of this year.

The Roku survey segregates the television market as follows: 43% of homes still have traditional cable TV. Another 25% are cord-shavers and still have traditional cable TV but have downsized to a lower-cost video package. 25% of the market are now cord-cutters, and 7% of the market never had traditional cable TV.

Probably the most interesting statistic is that one-fourth of the market consists of cord-shavers who have reduced their traditional programming packages. It’s been clear that cord-shaving has been happening, but I’ve never seen it quantified before. The big cable companies never mention cord-shaving when reporting cable TV subscribers. The magnitude of the number of households that have trimmed back to lower-cost programming packages explains why the paid subscriptions to cable networks is dropping far faster than the drop in cable customers.

80% of cord-cutters say that they are satisfied with their decision to end their subscription to traditional TV. Two-thirds of cord-cutters say they wish they had cut the cord sooner.

Lack of sports is driving some cord-cutting during the pandemic, and 28% of cord-cutters say that lack of sports is their number one reason for cutting the cord. 17% of cord-cutters (or 4% of the whole video market) say they will consider returning to traditional TV when sports return to the air full time. 31% of cord-cutters say they will pursue a sports streaming service when sports returns.

The number one reason cited for cutting the cord is cost savings, and many of those surveyed say they were driven to this decision due to a change in household income due to the pandemic. The average Roku user said that they are saving $75 per month with cord-cutting. As a household that has cut the cord, I find that number a little hard to believe – but it’s what they reported in the survey. My consulting firm does surveys and we’ve learned to always be leery when households cite numbers of any kind; in this case, it would be natural for many homes to exaggerate their savings as a way to justify cutting the cord. I’m sure some homes have saved $75, but that seems like a high average and it doesn’t take more than a few subscriptions to online video services to eat into that savings.

Cord-cutters are watching more free ad-supported content as a way to control costs. 42% of cord-cutting households said that free content or extended free subscriptions to streaming services helped to convince them to cut the cord.

45% of the households in the cord-shaver category say they are likely to cancel traditional TV in the next six months. Every survey about cable TV I’ve seen for the last five years has included substantial numbers of homes that say they are about to drop cable TV – but then don’t. But this statistic is a lot higher than I’ve ever seen and indicates a lot of households are thinking about cutting the cord. It’s often a complicated decision for a home with multiple family members to finally cut the cord.

The pandemic makes it harder to discern long-term trends. This survey supports the industry belief that a lot of homes continue to drop traditional TV packages. But the pandemic provides several good reasons to drop a cable subscription that won’t be permanent. Sports will eventually come back to TV and sports fans are going to subscribe. As the economy rebounds, people will get back to work – it’s an easier decision to cut a $100 per month cable subscription when one or more people in a home are unemployed. The pandemic has also killed the creation of new programming content, and many cable subscribers only pay in order to watch the latest versions of their favorite shows. I’ve read that it might take more than a year after the pandemic ends to see a fresh supply of new content.

It will take time to see if an improved economy reverses any of the cord-cutting trends. For now, any company offering cable TV is in for a rough ride. It’s hard to see any positive news from the results of this survey for programmers or cable companies.

Telemedicine and Broadband Access

North Carolina’s Broadband Infrastructure Office and the NC Department of Health and Human Services published a 77-page report that compares the state of healthcare and broadband access in Western North Carolina. I expect to soon see similar reports from around the country as States are taking a hard look at broadband access and related issues.

Western North Carolina is in Appalachia and has higher rates of poverty than the rest of the state. Many of the counties in the region had economies driven by coal extraction and related supply chains. These areas were already seeing economic devastation before the pandemic.

As might be expected, the report shows that deaths from diabetes, stroke, heart disease, opioid use, COPD, and unintentional injuries are higher in western North Carolina than in the state’s metropolitan areas. The region has far fewer doctors per 10,000 population than the rest of the state.

There had already been a start for bringing telemedicine to the region. For example, Mission Health, a large health care provider in Asheville had already been working with rural hospitals in a few western counties to bring access to specialists. But like everywhere else in the country, the need for telemedicine has exploded since the advent of the pandemic.

As might be expected, broadband access is low in many of these counties. Western North Carolina is like the rest of Appalachia, with hilly and mountainous terrain, a lot of woods, narrow and winding country roads, and scattered rural populations. The counties in the region have already identified the lack of broadband as a major problem before the pandemic and have been taking steps to try to attract ISPs to the region.

This study is the first step attempt in correlating broadband access to the availability of health care. One of the first steps taken in the study was to equate the broadband adoption rate in counties to the degree to which a county has higher death rates than the rest of the state. Interestingly, there were no counties with high broadband adoptions that rated below average for health statistics. However, only 3 or the 20 counties were rated as having high broadband adoption rates.

The study surveyed what the report called safety net sites – locations that provide health care to low-income people. They found that 70% of these locations were already using telemedicine before the pandemic. However, most of these health care providers said they were underutilizing telemedicine. The study showed that many of the health care facilities don’t have an affordable or reliable broadband connection, making it hard for them to reliably conduct telemedicine.

The State Broadband Office had identified almost 72,000 homes in these rural counties that don’t have access to broadband at home, meaning there is no ISP able to provide a broadband connection capable of delivering telemedicine. Telemedicine platforms differ, but I’ve been told that most telemedicine connections require both a download and an upload connection of between 3 Mbps and 4 Mbps. Anybody living in a rural area knows that an upload speed of that magnitude is a rarity.

The study also showed that even where broadband is available that 17 of the 20 counties have lower than average rates of computer ownership and broadband adoption. The study correlates this with the level of poverty, which is also lower in these areas than average for the State.

This study is an important step in understanding the broadband gaps in Western North Carolina, and telemedicine is only one of the many ways that lack of broadband is hurting the region. If any good has come out of the pandemic, it’s that state government has turned its attention to the huge problems caused by poor broadband in rural areas. Hopefully, this will translate into finding broadband solutions – which is going to be a huge challenge in Appalachia.

A New Rural Broadband Product?

Verizon announced a new wireless data product that raises a few questions for me. Verizon announced the ‘LTE Home Internet’ product on this web page. The product is easily explained. Verizon will be delivering unlimited data using the cellular 4G LTE network. Customers must buy a receiver from Verizon for $240, but the company is offering a $10 discount for 24-months which returns the cost of the box over two years. The product is $40 per month for a household that is buying a Verizon cellular product that costs at least $30 per month. Non-Verizon wireless customers pay $60 per month. There is free tech support for setup issues for 30-days, implying that tech support will entail a fee after that.

Verizon touts the product as delivering 25 Mbps download speeds, with bursts as high as 50 Mbps. Verizon is launching the product in three markets – Savannah, GA, Springfield, MO, and the Tri-cities area at the area near the borders of Tennessee, Virginia, and Kentucky.

The first question raised is if this product is intended to replace Verizon’s rural hotspot product, marketed as Verizon Jetpack. The Verizon announcement says, “Verizon will expand home Internet access to customers outside the Fios and 5G Home footprints, expanding home connectivity options to rural areas.”, which implies that this is a replacement for the current rural 4G hotspot product. If so, this would be a drastic repricing for rural LTE broadband.

The Jetpack hotspot is widely used in rural America where there are no other broadband alternatives. From what I can see, the Verizon hotspot is the most expensive broadband in the country and is billed at data rates similar to normal cellular plans. The Jetpack product has four available pricing tiers based upon the monthly data allowance. The 10 GB plan is $60, the 20 GB plan is $90, the 30 GB plan is $120, and the 40 GB plan is $150. The real price killer is that Verizon bills additional gigabytes at $10 each. I’ve talked to rural households that spend $500 or per month or more for the hotspot plan.

The release of the new product caught the industry by surprise and there was little or no buzz that this was coming. The big question that those living in rural America will have is if Verizon will offer this as an alternative to the Jetpack product. Is Verizon planning to move customers from plans that cost hundreds of dollars per month to a plan that offers unlimited data for $40 to $60? If so, then this is great news for rural America.

My second question concerns data speeds. Verizon advertises the existing Jetpack product as having from 5 to 12 Mbps download and 2 to 5 Mbps upload. However, the current plan comes with a warning that the product only works where Verizon has a ‘strong’ data signal. I’ve talked to a number of households that say that the Jetpack product is only delivering a few Mbps. Rural LTE data speeds are reliant upon two factors – how close a customer is to a cell tower, and the underlying strength of the broadband feeding the tower.

I wonder if the new product will be any faster? There is a chance that it can be faster if the new device utilizes more frequency bands than the old hotspot receiver. But cellular speeds, in general, get weaker with distance from a cellular tower, and folks that are more than a mile or two from a tower are not likely to get the touted 25 Mbps speeds on the new product.

The cynic in me suspects that Verizon will only activate this product near markets where they have faster broadband products. This would be a good fill-in product for low-bandwidth homes in neighborhoods served by FiOS or the new fiber-to-the-curb FWA product. This is not a bad broadband product for a home that only reads emails and watches a single stream of video – but this product would bog down quickly if used to support multiple simultaneous users.

I doubt that the average urban broadband customer appreciates the misery of homes using the Jetpack hotspot. Data use is metered and it cost $10 of broadband to watch a movie. Families with kids using the hotspot have a constant fight to keep them off the Internet. I hope my gut feeling is wrong and that Verizon will introduce this everywhere and toss out the hotspot product. Even if this product doesn’t bring faster data speeds to rural homes, the pricing and the ability to use unlimited data would be a welcome relief to homes using the Jetpack hotspot. It’s possible that this product is Verizon’s response to T-Mobile’s promised rural 5G product, but we’ll have to wait to see where this is made available before getting too excited about it.

Time to Stop Talking about Unserved and Underserved

I work with communities all of the time that want to know if they are unserved or underserved by broadband. I’ve started to tell them to toss away those two terms, which are not a good way to think about broadband today.

The first time I remember the use of these two terms was as part of the 2009 grant program created by the American Recovery & Reinvestment Act of 2009. The language that created those grants included language from Congress that defined the two terms. In that grant program, unserved meant any home or business that has a broadband speed of less than 10/1 Mbps. Underserved was defined as homes having speeds above 10/1 Mbps but slower than 25/3 Mbps.

As far as I can tell, these terms have never been defined outside of broadband grant programs. However, the terms began to be widely used when talking about broadband availability. A decade ago, communities all wanted to know if they were unserved or underserved.

The terms began to show up in other grant programs after 2009. For example, the FCC’s CAF II grant program in 2015 gave money to the largest telephone companies in the country and funded ‘unserved’ locations that had speeds less than 10/1 Mbps.

The same definition was used in the ReConnect grants created by Congress in 2018 and 2019. Those grants made money available to bring better broadband to areas that had to be at least 90% unserved, using the 10/1 Mbps definition.

The biggest FCC grant program of 2020 has scrapped the old definition of these terms. This $20.4 billion Rural Digital Opportunity Fund (RDOF) grant program is being made eligible to Census blocks that are “entirely unserved by voice and with broadband speeds of at least 25/3 Mbps”. That seemingly has redefined unserved to now mean 25/3 Mbps or slower broadband – at least for purposes of this federal grant program.

There are also states that have defined the two terms differently. For example, following is the official definition of broadband in Minnesota that is used when awarding broadband grants in the state:

An unserved area is an area of Minnesota in which households or businesses lack access to wire-line broadband service at speeds that meet the FCC threshold of 25 megabits per second download and 3 megabits per second upload. An underserved area is an area of Minnesota in which households or businesses do receive service at or above the FCC threshold but lack access to wire-line broadband service at speeds 100 megabits per second download and 20 megabits per second upload.

It must also be noted that there are states that define slower speeds as unserved. I’m aware of a few state broadband programs that still use 4/1 Mbps or 6/1 Mbps as the definition of unserved.

The main reason to scrap these terms is that they convey the idea that 25/3 Mbps broadband ought to be an acceptable target speeds for building new broadband. Urban America has moved far beyond the kinds of broadband speeds that are being discussed as acceptable for rural broadband. Cable companies now have minimum speeds that vary between 100 Mbps and 200 Mbps. Almost 18% of homes in the US now buy broadband provided over fiber. Cisco says the average achieved broadband speed in 2020 is in the range of 93 Mbps.

The time has come when we all need to refuse to talk about subsidizing broadband infrastructure that is obsolete before it’s constructed. We saw during the recent pandemic that homes need faster upload speeds in order to work or do schoolwork from home. We must refuse to accept new broadband construction that provides a 3 Mbps upload connection when something ten times faster than that would barely be acceptable.

Words have power, and the FCC still frames the national broadband discussions in terms of the ability to provide speeds of 25/3 Mbps. The FCC concentrated on 25/3 Mbps as the primary point of focus in its two recent FCC broadband reports to Congress. By sticking with discussions of 25/3 Mbps, the FCC is able to declare that a lot of the US has acceptable broadband. If the FCC used a more realistic definition of broadband, like the one used in Minnesota, then the many millions of homes that can’t buy 100/20 Mbps broadband would be properly defined as being underserved.

In the last few months, the FCC decided to allow slow technologies into the $16.4 billion RDOF grant program. For example, they’ve opened the door to telcos to bid to provide rural DSL that will supposedly offer 25/3 Mbps speeds. This is after the complete failure in the CAF II program where the big telcos largely failed to bring rural DSL speeds up to a paltry 10/1 Mbps.

It’s time to kill the terms unserved and underserved, and it’s time to stop defining connections of 10/1 Mbps or 25/3 Mbps as broadband. When urban residents can buy broadband with speeds of 100 Mbps or faster, a connection of 25/3 should not be referred to as broadband.

An Update on LEO Satellites

A lot of rural America continues to hope that low orbit satellite (LEO) service will provide a broadband alternative. It’s been a while since I’ve covered the status of the companies proposing to deploy constellations of satellites for providing broadband.

In March, OneWeb filed for Chapter 11 restructuring when it was clear that the company could not raise enough cash to continue the research and development of the satellite product. In July, a bankruptcy court in New York approved a $1 billion offer to take over the company filed jointly by the British Government and Bharti Airtel. Airtel is India’s largest cellular company. The restructured company will be owned with 45% stakes by Britain and Bharti Airtel, with the remaining 10% held by Softbank of Japan, the biggest original shareholder of OneWeb. Other earlier investors like the founders, Intelsat, Totalplay Telecommunications of Mexico, and Coca-Cola have been closed out of ownership by the transaction.

There is speculation that the British government purchased the company to create tech jobs in the country and that all R&D and manufacturing for OneWeb would immediately shift to England from Florida.

Of more concern for rural broadband is speculation that the mission of the company will change. Greg Wyler, the original CEO of the company had a vision of using the satellites to bring broadband to parts of the world that have no broadband. He chose a polar orbit for the satellites and was going to launch the business by serving Alaska and the northern territories of Canada like Nunavut. I’ve seen speculation that the revised company is likely to concentrate instead on wholesale connections to telcos and ISPs, such as providing backhaul for rural cell sites.

Elon Musk’s satellite venture StarLink was recently in the news when the company said it was going to raise ‘up to $1 billion’ to continue the development of the business. The company still has a long and expensive road to success. The company has raised over $3.5 billion to date before this latest raise, but a recent Bloomberg article estimates that the company will need to raise an additional $50 billion between now and 2033, which is when the company is projected to be cash-positive.

StarLink now has over 540 satellites in orbit, but the business plan calls for over 4,000. Keeping the constellation in place will be an ongoing challenge since the satellites have an estimated life of 5 to 6 years. Starlink will forever have to be launching new satellites to replace downed satellites.

The US government and the FCC seem to be in StarLink’s corner. The FCC is still evaluating if it will allow StarLink to participate in the upcoming RDOF grants auction in October. It would be incredibly unusual to award giant federal grants for a product that is still on the drawing board and for an ISP that hasn’t raised 10% of their needed funding.

StarLink recently made a very-public announcement that it was looking for beta customers – likely as a way to spur fundraising. Early Starlink customers will likely see blazingly fast speeds, which would happen for any broadband technology that could devote the bandwidth from one server to connect to one or two customers. The bandwidth delivered on a fully-subscribed satellite network will be far less – but that won’t stop the company from using a beta test to set unrealistic expectations of future satellite broadband speeds.

The last LEO player that is still active is Jeff Bezos venture that is still using the preliminary name of Project Kuiper. The FCC recently approved the licensing for Project Kuiper to move forward. Immediately following the FCC approval, Jeff Bezos announced that he will be investing $10 billion in the business. This ability to self-fund likely gives Project Kuiper an advantage over other competitors. It was reported that just for the month of July that Bezos’s net worth had climbed by $9 billion.  Funding is going to be a constant hurdle for the other two major competitors, but Project Kuiper might be the fastest to deploy if funding is not an issue.

The FCC approval pf Project Kuiper and the funding announcement by Bezos came at the same time that Starlink is seeking another round of financing and is trying to get into the FCC auction. It’s going to be interesting to see how the battle between two billionaires unfolds – my bet is on Amazon due to easy access to funding.

Is the Line Between Wireless and Wireline Blurring?

In the Bernstein Strategic Conference in May, Ronan Dunne, Verizon CEO and EVP for Verizon Consumer talked about his vision for the future of 5G. During that presentation, he made a statement that has been bugging me for weeks, so I finally had to write about it. He said that he can foresee a day when consumers will purchase home broadband in the same way that they buy wireless service today. He said that will happen because the line between the wireless and wireline business are blurring.

Dunne is talking about a future when 5G is ubiquitous and where people won’t perceive a difference between landline broadband and 5G broadband. In a term used by an economist, Dunne foresees a day when wireless broadband is a pure substitute for landline broadband – where a customer won’t perceive a functional difference between the two products.

Verizon offers several wireless products, so let’s talk about them individually. The predominant Verizon product that is in every market is cellular broadband. This uses cell sites to beam voice and data traffic to cellphones or other devices that are connected to a cellular data plan. Those cellular plans are incredibly stingy in terms of the amount of broadband that can be used in a month, with the unlimited plans offering a little more than 20 gigabytes of data before a user has to pay more or become restricted. The specifications for 5G set a goal of 100 Mbps for cellular broadband speeds within a decade. That kind of speed might be a substitute for landline broadband today from a speed perspective. But networks are not likely to achieve these speeds for at least five more years, and by then I think cable companies will be considering increase urban broadband speeds to something like 250 Mbps. I have to question if cellular broadband speeds can keep up with the speeds provided by landline connections.

Of more importance is that cellular speeds drop when entering a building. Anybody who has walked into a large building using their cellphone understands that cellular signals don’t perform as well indoors as outdoors. By the time I walk 100 feet into my neighborhood grocery store, I often have zero bars of data. While speeds don’t drop that drastically in most homes, when outdoor cellular speeds hit 100 Mbps, indoor speeds in most homes might hit half that number. With slower speeds and incredibly stingy data caps it’s hard to see cellular broadband as a pure substitute for a landline broadband connection.

I also don’t think that the gimmick product that Verizon and others are selling in urban city centers that offers gigabit speeds using millimeter wave spectrum is a landline substitute. The product requires closely spaced small cell sites fed by fiber – but the big gotcha is that the millimeter wave spectrum won’t penetrate a building and barely even make it through a pane of glass. This is an outdoor product for which I still struggle to understand a willing market. It’s certainly not a substitute for landline broadband, except perhaps for somebody who is always outdoors.

The newest wireless product is Verizon’s fixed wireless access (FWA) that beams a broadband signal into the home from a pole-mounted transmitter at the curb using millimeter wave spectrum. I have to suspect that this is the product Dunne is talking about. I would agree with him that this is a pure substitute for landline broadband. But that’s because this is just another variation of landline broadband. This technology has historically been referred to as fiber-to-the-curb. Verizon is using a wireless transmission instead of a fiber line for the last hundred feet to reach a home – but this technology requires building the same fiber into neighborhoods as fiber-to-the-home. This is not a wireless technology since 99% of the network is still comprised of fiber. Anybody using this service can walk to their curb to see the fiber that is carrying their broadband. This technology is a clear substitute for a landline fiber drop – but it’s not a wireless network other than for the last 100 feet to a home.

The other way to challenge Dunne’s vision is by comparing the volume of traffic used by landline and wireless networks. The vast majority of data traffic is still carried over wires and the gulf between the data carried by each technology is widening every year. Consider the following chart from Cisco from 2019 that shows the volumes of monthly data traffic in North America by type. This is expressed in exabytes (one billion gigabytes).

Monthly Exabytes 2017 2018 2019 2020 2021 2022
Homes 35 43 53 64 75 90
Cellular 1.3 1.8 2.5 3.4 4.5 5.9
Business 6.5 8.3 10.3 12.8 15.5 18.5
Total 43 53 66 80 95 114

Both home and business broadband are carried on wires. In 2020, only a little more than 4% of all of the data traffic in North America is carried wirelessly. For wireless technology to be a pure substitute for wireline data, wireless networks would have to be capable of carrying a much bigger share of data – many times what they carry today. The laws of physics argue against that, particularly since landline data usage is growing at an exponential rate. It’s hard to envision wireless networks in our lifetime that can handle the same volumes of data as fiber-based landline networks.

This is not intended as a major criticism of what Dunne said. The country will be better off if Verizon offers a competitive alternative to the cable companies. However, Verizon is like the other cellular companies and can’t talk about 5G without overstating the potential. I know has to keep hyping 5G for Wall Street and I sympathize with that need. But we are still very far from a day when the average household will view landline and wireless data to be pure substitutes.

Customers Still Flock to Promotional Rates

FierceVideo and others recently reported on a survey done in June by the research firm Cowen that looked at consumer use of promotional rates.

Cowen found that 20% of big ISP subscribers are on Internet plans that have promotional rates that will expire within the next 12 months. Another 13% of subscribers are on promotional plans that will expire in a time frame longer than 12 months. Surprisingly, 10% of subscribers have price-for-life guarantees. This leaves just 57% of subscribers paying full price for ISP services.

Promotional pricing is a sensitive topic for the industry and none of the big cable companies or telcos disclose the volume or amounts of discounts they give to customers. The big ISPs are all under a lot of pressure from Wall Street, and one of the key metrics used by analysts to track the big companies is ARPU – average revenue per user. ISPs have hard decisions to make. Giving too many discounts can kill ARPU, but not offering discounts can lose customers and revenues.

Some big ISPs have been working to curtail promotional pricing. AT&T has lost nearly three million video customers in the last year and claims that the losses mostly are due to tightening the promotional pricing that was given in the past by DirecTV. It’s also been reported that Charter has been tightening its policies on promotional prices, and in particular was ending a huge volume of promotional pricing they inherited through the acquisition of Time Warner Cable.

The Cowen report highlighted the difference in discount philosophy varies by ISP. For example, the report said that 45% of Altice customers have a promotional package, Comcast has 42%, and Charter is at 32%.

The big ISPs dole out promotional discounts in a few different ways. All of the incumbent ISPs offer low prices on the web to attract new customers. These new customer discounts generally last for 12 to 24 months before customers are moved to normal pricing. The other big category of promotional discounts is discounts that are negotiated with customers, often when customers threaten to leave an ISP.

The Cowen study confirmed something that we’ve always seen in the market. The promotional prices tend to go to younger subscribers, and older customers tend to pay full price for services. It takes real effort to either change ISPs or to renegotiate pricing every year or two, and only consumers willing to go through that hassle end up with a repetitive series of promotional deals.

The statistic that surprised me was that 10% of respondents in the survey said they had lifetime rates. ISPs have been somewhat leery of using the ‘lifetime rate’ words, but over the years as ISPs increased speeds and prices on their networks they have often allowed customers to stick with slower and less expensive broadband – generally with the caveat that a customer with a grandfathered plan can make no changes without being moved to newer pricing. In my mind, there is a significant difference between grandfathering an existing plan that offers slower speeds than other customers compared to new lifetime sales promotions that offer such deals to new customers. One of the biggest advantages to the ISPs of grandfathered plans is that customers keep these plans for years, meaning no churn.

Small ISPs struggle with promotional rates. Some small ISPs that still offer video offer guaranteed bundled rates for customers who buy cable TV. But I know a number of small ISPs that have ceased offering bundled discounts since the margins on cable TV are too small to afford them.

Small ISPs also generally don’t like the hassle of always having to negotiate rates with customers seeking a discount. Negotiating with customers changes the culture in a call center and adds a lot of pressure to customer service reps – and is probably the number one reason why the public dislikes big ISP customer service.

Many small ISPs have also given up on the idea of having residential service contracts. It’s a major pain to collect from somebody who breaks a contract and drops service. Most of the small ISPs I know feel that their quality of service is superior to the competition and they don’t want to fight to keep unhappy customers.

Bandwidth Needed to Work from Home

The pandemic made it clear that the millions of homes with no broadband or poor broadband were cut off from taking the office or the school home. But the pandemic also showed many additional millions of homes that their current ISP connection isn’t up to snuff for working or doing schoolwork from home. Families often found that multiple adults and students couldn’t share the bandwidth at the same time.

The simplest explanation for this is that homes were suddenly expected to connect to a school or work servers, use new services like Zoom, or make telemedicine connections to talk to doctors. These new requirements have significantly different bandwidth needs when a home’s big bandwidth need was watching multiple video streams at the same time.  Consider the following bandwidth needs listed by Zoom:

Zoom says that a home should have a 2 Mbps connection, both upload and download to sustain a Zoom session between just two people. The amount of download bandwidth increases with each person connected to the call, meaning Zoom recommends 6 Mbps download for a meeting with three other people.

Telemedicine connections tend to be even larger than this and also require the simultaneous use of both upload and download bandwidth. Connections to work and schools servers vary in size depending upon the specific software being used, but the VPNs from these connections are typically as large or larger than the requirements for the Zoom.

Straight math shows fairly large requirements if three or four people are trying to do make these same kinds of 2-way simultaneous connections at the same time. But houses are also using traditional bandwidth during the pandemic like watching video, gaming, web browsing, and downloading large work files.

The simplistic way to look at bandwidth needs is to add up the various uses. For instance, if four people in a home wanted to have a Zoom conversation with another person the home would need a simultaneous connection of 8 Mbps both up and down. But bandwidth use in a house is not that simple, and a lot of other factors contribute to the quality of bandwidth connections within a home. Consider all of the following:

  • WiFi Collisions. WiFi networks can be extremely inefficient when multiple people are trying to use the same WiFi channels at the same time. Today’s version of WiFi only has a few channels to choose from, and so the multiple connections on the WiFi network interfere with each other. It’s not unusual for the WiFi network to add a 20% to 30% overhead, meaning that collisions of WiFi signals effectively waste usable bandwidth. A lot of this problem is going to be fixed with WiFi 6 and 6 GHz bandwidth which together will add a lot of new channels inside the home.
  • Lack of Quality of Service (QoS). Home broadband networks don’t provide quality of service, which means that homes are unable to prioritize data streams. If you were able to prioritize a school connection, then any problems inside the network would affect other connections first and would maintain a steady connection to a school. Without QoS, a degraded bandwidth signal is likely to affect everybody using the Internet. This is easily demonstrated if somebody in a home tries to upload a giant data file while somebody else is using Zoom – the Zoom connection can easily drop temporarily below the needed bandwidth threshold and either freeze or drop the connection.
  • Share Neighborhood Bandwidth. Unfortunately, a home using DSL or cable modems doesn’t only have to worry about how other in the home are using the bandwidth, because these services used shared networks within neighborhoods, and as the demand needs for the whole neighborhood increase, the quality of the bandwidth available to everybody degrades.
  • Physical Issues. ISPs don’t want to talk about it, but events like drop wires swinging in the wind can affect a DSL or cable modem connection. Cable broadband networks are also susceptible to radio interference – your connection will get a little worse when your neighbor is operating a blender or microwave oven.
  • ISP Limitations. All bandwidth is not the same. For example, the upload bandwidth in a cable company network uses the worse spectrum inside the cable network – the part that is most susceptible to interference. This never mattered in the past when everybody cared about download bandwidth, but an interference-laden 10 Mbps upload stream is not going to deliver a reliable 10 Mbps connection. There are a half dozen similar limitations that ISPs never talk about that affect available bandwidth.

The average home experiencing problems when working at home during the pandemic is unlikely to be able to fully diagnose the reasons for the poor bandwidth. It is fairly obvious if you are having problems with having multiple zoom connections if the home upload speed isn’t fast enough to accommodate all of the connections. But beyond the lack of broadband capacity, it is not easy for a homeowner to understand any other local problems affecting their broadband experience. The easiest fix for home broadband problems is for an ISP to offer and deliver faster speed, since excess capacity can overcome many of the other problems that might be plaguing a given home.