Tag Archives: Amazon

Is Satellite Cellular a Competitive Threat?

It seems that one of the hottest topics in the industry is Direct-to-Device (D2D) cellular service that can bypass terrestrial cell towers and connect cellphones directly to satellites. I’ve already seen claims being made that D2D technology will solve the rural cellular issue since everybody will have cell coverage. Today’s blog looks at some of the reality behind these claims.

There are a lot of players suddenly in this space. Not surprisingly, SpaceX / Starlink already has hundreds of satellites in orbit and plans for 15,000. AST SpaceMobile is partnering with AT&T and Verizon. Globalstar originally pioneered Apple’s SOS emergency test service and was recently acquired by Amazon, which recently applied for licenses for 5,105 satellites for D2D. Lynk Global has partnered with more than 45 cellular companies worldwide. EchoStar and Viasat have also announced plans to enter the market.

Basic Technology. The satellites will circle at heights between 200 and 550 miles above Earth and will be traveling at 17,000 miles per hour. At least for now, nobody has figured out how to hand off calls between a satellite and terrestrial cellular network. There are some limitations for D2D that are largely defined by physics.

  • Can’t Work Indoors. Your cellphone transmits at under 0.5 watts of power – the maximum power level set by the FCC. Cell signals lose from 50% to 90% of the signal strength when passing through a wall. While that usually is enough power to reach a cell tower located within 2-3 miles of your house, it’s not nearly enough power to reach a satellite.
  • Receiver Limits. Starlink has been achieving data speeds up to 300 Mbps download for home broadband. Expect speeds to be far less for a D2D due to the much smaller antenna surface area in a cellphone compared to a Starlink receiver dish. In general, the larger the size of the antenna, the more bits that can be gathered. Receivers for Starlink range from 11.75 X 15.7 inches for the Starlink Mini to 22.7 X 20.1 for the higher-performance receiver. The typical cellphone has a receiver of around 6.5 X 7 millimeters, meaning that even the Starlink Mini has 2,600 times more surface area for receiving a signal from the satellite than a typical cellphone.
  • Area Speed Coverage Limits. The planned coverage area for the next generation of D2D Starlink satellite is around 45 square miles. The newer satellites will be using massive adjustable antenna beams along with 40 to 65 MHz channels. This means that a single data stream from a satellite will have an overall data capacity between 40 and 150 Mbps download, depending on the spectrum being used. Starlink is planning on 16 simultaneous full-power beams from the next generation D2D satellites. The challenge is that this spectrum must be shared among all users inside a beam footprint. That means a maximum of 2.4 Gbps of bandwidth to share in a 45 square mile area – a minuscule amount of bandwidth compared to what can be delivered by cell towers serving the same area. This is compounded by a limitation on the simultaneous number of users that can be connected at the same time.

What does this all mean from a practical perspective? The D2D satellites will be able to connect to a large number of people who are outdoors, within the overall limitation of the number of simultaneous people that can be connected at one time. We won’t know until somebody shows us what this will mean for connecting a lot of people in a small area, like at a rural rock concert.

Early trials have shown that D2D has some capacity for making a connection to a moving vehicle. This is never going to be as guaranteed as when standing outdoors, and it will likely depend on the type of vehicle and the speed being driven.

D2D is not going to work directly indoors. It’s possible that somebody will develop an outdoor receiver that can bounce a signal to go in through a window. It would make sense to do this in conjunction with something like a Starlink receiver.

It’s hard to envision D2D as a standalone cellular service other than possibly for the rare person who spends most of their time outdoors.

To be fair to the industry, at least for now, the big D2D providers aren’t touting it as a standalone cellular service. They are talking about this being an add-on to normal cellular plans. This doesn’t make much sense in urban areas where normal cell coverage is already ubiquitous and strong. But this could be an interesting additive to a cell plan for folks who either live or travel a lot in rural areas. It’s a cool technology, but a lot of the excitement could be empty hype, which seems to be the norm for most new technology products.

There is only one statistic that is really going to matter for D2D, which is the number of people willing to pony up a monthly fee to use it. The one thing going for the new technology is that there is a lot more serious demand for any kind of cellular coverage in the third world than there is in the rural U.S. Maybe that’s the real market they are shooting for.

SpaceX Wants to Handle the World’s Broadband

In a recent request filed with the FCC, Starlink asked to increase the size of its broadband satellite fleet to 100,000. This coincided with the FCC’s announcement that it wants to ease the process for approving new satellites, so I have to think the idea will sail through. There was one extraordinary statement in this filing that is the topic of today’s blog. SpaceX said it has the goal of handling the majority of the world’s Internet traffic. Elon Musk has always been one for huge hyperbole, so statements like this are not surprising. I want to examine a few reasons why that goal is ridiculous.

No carrier handles a large percentage of the world’s internet traffic today. According to statistics shown on Cloudflare Radar, the company with the biggest percentage of worldwide traffic today is Amazon, with 3.5% of all traffic. The companies that carry more than 2% of worldwide traffic include Cloudflare (2.2%), Comcast (2.5%), Amazon (2.4%), Microsoft (2.2%), Reliance/Jio (2.1%), and AT&T (2.1%). SpaceX is on the worldwide list at number 18 today, already carrying 0.7% of the world’s Internet traffic. SpaceX would have to grow by five times to catch Amazon. But that would still be a long way from carrying a majority of the world’s traffic, and it’s hard to imagine any carrier growing to carry even 10% of the world’s traffic.

Starlink has built an impressive ISP business by gaining 12 million broadband customers. Worldwide, there are over 6 billion Internet users, about 74% of the world’s population. If you assume an average of 2.5 people per household, Starlink today represents 0.5% of all residential Internet users. But residential broadband, as large as it is, is still only a fraction of the world’s Internet traffic. Again, according to Cloudflare, bots are currently making 56.8% of all web connections.

Starlink largely has a near-monopoly for satellite broadband today, but that’s going to change over the next five years. There are major plans by others to compete with Starlink, including Amazon, China’s Guowang, Shanghai’s Quinfan, Blue Origin’s TeraWave, and Eutelsat’s OneWeb. Competition will dilute Starlink’s share of the satellite market.

Love him or hate him, Elon Musk has aligned himself with politics, which is unusual for somebody operating a giant retail business. There are already many countries that ban Starlink, and this list could grow. Even in countries where Starlink is available, there are a lot of people who would never buy service from an Elon Musk-backed company.

I think the biggest reason is that there is a lot more risk associated with trusting your broadband to satellites. A massive 1,000-year solar flare could knock a lot of satellites out of commission. There is a real threat of a spreading cloud of space debris killing a lot of satellites, as predicted by the Kessler syndrome. There is a chance that satellites could be knocked out deliberately during a war or by a rogue nation.

Fiber networks face no similar worldwide threats. Large users of broadband are never going to trust their traffic to satellite. Don’t expect to see data centers, universities, governments, or large corporations favor satellite connections over fiber.

I have to admit that I like Elon Musk’s brashness, but his claims are always overboard. I think he has a better chance of establishing his colony on Mars than of capturing the majority of the world’s Internet traffic.

Market Consolidation Continues

It looks like industry mergers and acquisition activity is in high gear lately. It’s hard to remember a week when there wasn’t a press release about upcoming M&A activity in the telecom sector, and I have been writing a similar blog every six months. Following is some of the most recent activity.

In the ISP Space. T-Mobile announced it entered two joint ventures to acquire 50% of three U.S. fiber businesses – GoNetspeed, Greenlight Networks, and i3 Broadband. T-Mobile seems to be gobbling up last-mile fiber properties all over the country.

TDS Telecom announced plans to buy Granite State Communications, a telco in New Hampshire with more than 11,000 service addresses.

Truvista Fiber is buying the municipal fiber network from the City of Commerce, Georgia, with plans to expand to reach residential customers.

Middle-Mile / Networks. Zayo just closed on the $4.25 billion acquisition of the fiber assets of Crown Castle Fiber. This adds 90,000 miles of fiber to Zayo’s U.S. network.

The managed service providers Spectrotel and Airespring announced a merger to become more competitive in serving business customers.

GCI in Alaska is acquiring Q Gateway Intermediate Holdings (Quintillion), a fiber infrastructure provider in the state. The purchase brings 1,800 miles of subsea and terrestrial fiber, along with active construction on additional routes.

Lumen is buying the cloud network company Alkira for $475 million. This brings expertise in AI programmable networking. Lumen has obviously decided to beef up its enterprise business now that it recently closed on the sale of last-mile fiber customers to AT&T.

Vendors. Inseego, a wireless edge vendor, is buying the FWA business line from Nokia.

Render Networks is acquiring mPower, a company that makes management tools for electric and water utilities.

Satellite. Amazon announced plans to purchase Globalstar for $11 billion. This jump-starts Amazon’s entry into the direct-to-device market.

The Mother of all Merger Rumors. In what would be the biggest telecom merger ever, Fierce Networks had a story about analysts at New Street Research who are speculating that a merger between Comcast and Charter makes a lot of sense. They said that Charter is still open to further acquisitions after it closes on the merger with Cox Communications. The article even speculated on Charter being an acquisition target for T-Mobile or SpaceX.

We can’t forget the three big ISP mergers of Charter/Cox, AT&T/Lumen Fiber, and Verizon/Frontier. The biggest ISPs are suddenly getting a lot larger.

Amazon’s Huge IoT Network

In a recent blog post, Amazon invited developers to test drive its gigantic IoT network. This network has been labeled as Sidewalk and was created by tying together all of Amazon’s wireless devices like Amazon Echos and Ring cameras.

Amazon claims this huge wireless network now covers 90% of U.S. households. Amazon created the network by transmitting Bluetooth and 900 MHz LoRa signals from its various devices. This network provides a benefit to Amazon because it can detect and track its own devices separate from anything a homeowner might do with WiFi.

But Amazon has intended for years to monetize this network, and this announcement begins that process. This network has been under-the-radar until now, and most homeowners have no idea that their Amazon devices can connect and communicate with other devices outside the home. Amazon swears that the IoT connection between devices is separate from anything happening inside the house using WiFi – that the IoT network is a fully separate network.

Anyplace where there are more than a few Amazon devices, the network should be robust. The 900 MHz spectrum adds a lot of distance to the signals, and it’s a frequency that does a good job of penetrating obstacles like homes and trees.

Amazon believes that this network can be used by IoT device makers to improve the performance of IoT devices in a neighborhood – things like smart thermostats, appliance sensors, and smart door locks. Such devices use only a small amount of bandwidth but are reliant on the home broadband network being operational to work. Amazon’s vision with this network is that your smart door lock will still work even when your home WiFi isn’t working.

By making the network available to others, Amazon can unleash developers to create new types of wireless devices. For example, it’s always been a challenge to use outdoor sensors since WiFi signals outside of homes is weak and inconsistent. It’s not hard to imagine a whole new array of sensors enabled by the Sidewalk network. Picture a motion detector on a shed door or a leak detector on outdoor faucets. With this network, vendors can now manufacture such devices with the knowledge that most homes will be able to make the needed wireless connection.

This also holds a lot of promise for municipal and business sensors. This is a low-cost way to communicate with smart city or other sensors. This would enable, for the first time, the deployment of environment sensors anywhere within range of the Sidewalk network.

This is another interesting venture by Amazon. At least in the U.S., this is a lower-cost solution than trying to connect to IoT devices by satellite. The only cost of building this network for Amazon was adding the wireless capability to its devices – mere pennies when deployed across millions of devices. But interestingly, Amazon will also have a satellite network starting in 2025 that can fill in the gaps where the Sidewalk network can’t reach.

Amazon says that it has already made deals to test the network with companies like Netvox, OnAsset, and Primax. Now that manufacturers know this network exists and is available, this ought to open up a wide range of new IoT devices that are not reliant only on WiFi. This might finally be the network that enables the original promise of IoT of a world with sensors everywhere, keeping tabs on the environment around us.

A Comcast Product for Cord-cutters

It’s been interesting to watch how the big cable companies have been trying to battle cord-cutting. Comcast has had a product for a while that is aimed directly at cord-cutters.  It’s labeled as Flex and is a video streaming service that is only available to Comcast broadband customers who are not buying a Comcast TV product.

Comcast hoped that Flex would be a direct competitor to Roku, Amazon Firestick, and Google Chrome. The Flex product clearly wants to keep customers who cut the cord inside the Comcast umbrella.

The product delivers 10,000 programs including Comcast content and content from other free online services like Pluto, Xumo, and Tubi. Content comes with commercials. For now, Comcast is also throwing in Comcast’s paid service Peacock for free. The Flex platform also gives customers an easy portal to watch Netflix, Amazon Prime, HBO, and Hulu.

Flex is certainly price right and currently is free for Xfinity broadband customers. When first introduced, Comcast was charging $4.95. Flex still requires the Comcast settop box and remote. I’m guessing the price came down when Comcast found few buyers willing to rent a box to get free content.

There is a big difference between Flex and competitors like Firestick or Roku in that customers can only use the apps on the platform that Comcast has installed. No additional apps for video or music services can be added to the app. This is probably the biggest disadvantage of Flex in that people are using a lot of different video apps online. I have an Amazon Firestick and it will let me add any online video app regardless of whether the app provider has a deal with Amazon.

When Comcast first introduced the Flex product, I thought the company wanted to become another superbundler like Amazon. Amazon allows customers to buy a subscription to a huge array of different online apps, and I assume Amazon gets a slice of revenue for delivering customers to partner video platforms. There are many video services for which Amazon has become the primary marketing channel. Amazon even suggests content that requires a subscription to the partner apps. The superbundler concept is likely profitable. Amazon has to be doing well taking a small slice of the revenue stream from nearly one hundred other platforms.

Amazon’s made it clear a few years ago that it wanted to become the one-stop shop for online video content, and Amazon has bundled together far more content than anybody else. But in the last year, we’ve seen the rise of proprietary platforms from CBS, Apple, Disney, and others including Comcast’s Peacock that won’t cross-market with Amazon and others. It’s not looking like any one bundler is going to be able to pull together a giant percentage of online video content.

It’s less clear how Comcast intends to benefit from Flex. I assume Comcast gets a share of ad revenues generated on platforms like Pluto. But there doesn’t seem to any other major benefit to the company for operating the Flex program, particularly since they are providing the settop box to Flex customers for free. The plan probably made financial sense at a monthly $4.95 rate, but it’s hard to see the long-term benefit to Comcast of offering a free service. Perhaps the one big benefit to Comcast is that the settop box used for Flex can also be used to control smart home and other Comcast products. Perhaps the company is using Flex to draw in customers for these other products.

Comcast has one big advantage over anybody else in the industry in that every Flex customer is already a Comcast broadband customer. That should mean that Comcast has little incremental cost for delivering the free content offered by Flex. It’s easy to forget that Netflix and all of the other online providers must maintain an expensive network to enable them to disseminate video content.

The Flex product is somewhat symbolic of the attempt for industry players to somehow be relevant in the online video market. The product doesn’t drive direct revenue for Comcast even though the company must provide a settop box. The platform is proprietary, which seems to be the new norm for video platforms. It’s one more of the many confusing choices faced by cord-cutters.

Amazon Joins the Broadband Space Race

I wrote a blog just a few weeks ago talking about how OneWeb had fully leaped into the race to place broadband satellites by launching a few test satellites and also by raising a few more billion dollars to fund the venture.

It’s been rumored for several years that Amazon was also interested in the idea, but their plans have been under wraps. It just came to light that Amazon has taken the first public steps and had the FCC file paperwork with the International Telecommunications Union to make notice of Amazon’s intent to launch satellites.

Amazon filed with the FCC under the name of Kuiper Systems LLC. Space fans will recognize the corporate name as a reference to the Kuiper belt, which is the area of the solar system past Neptune that is believed to contain numerous comets, asteroids and other small objects made largely of ice.

Amazon has big plans and the ITU filing said the company wants to launch a constellation of 3,236 satellites in low earth orbit. That’s 784 satellites in orbit at 367 miles above the earth, 1,296 in orbit at 379 miles, and 1,156 in orbit at 391 miles. Added to the other companies that are talking about getting into the business that’s now more than 10,000 planned satellites.

We know that Jeff Bezos is serious about space. He owns a rocket business, Blue Origins, that is developing an orbital-class rocket called the New Glenn. That company already has some future contracts to make private launches for OneWeb and Telesat. Amazon also recently launched a cloud computing service knows as AWS Ground Station that is intended to provide communications data links between earth and object in outer space. We also found out recently that Bezos kept 100% control of Blue Origins as part of his divorce settlement.

None of the low-orbit satellite ventures have talked about broadband speeds, prices or customer penetration goals. The only one making any announcement was SpaceX who said that his Starlink satellites would be capable of making a gigabit connection to earth. But that’s a far cry from a realistic estimate of a broadband product and is the satellite version of the Sprint cellphone test that showed that millimeter wave spectrum could deliver gigabit speeds to a cellphone. It can be done but is incredibly hard and would involve synching big data pipes from multiple satellites to a single customer.

We got another clue recently when OneWeb asked the FCC for permission to eventually create 1 million links to earth-based receivers, meaning customers. That puts some perspective on the satellites and shows that they are not trying to bring broadband to every rural customer. But still, one million satellite connections would represent about 10% of the rural homes in the US that don’t have broadband today. If that’s their US goal it automatically tells me that prices will likely be high.

NASA and others in charge of space policy have also started talking recently about the potential dangers from so many objects in orbit. We don’t know the size of the Amazon satellites yet. But Elon Musk said his satellites would range in size from a refrigerator down to some that are not larger than a football. NASA is worried about collisions between manned space flights with satellites and space debris.

Amazon is still early in the process. They haven’t yet filed a formal proposal to the FCC discussing their technology and plans. They are several years behind OneWeb and Starlink in terms of getting a test satellite into orbit. But an Amazon space venture has the built-in advantage of being able to advertise a satellite broadband product on the Amazon website where the vast majority of Americans routinely shop. I can envision Amazon measuring the broadband speed of a customer connected to the Amazon website and popping up an offer to buy faster broadband.

It’s absolutely impossible to predict the impact these various satellite companies will have on US broadband. A lot of their impact is going to depend upon the speeds and prices they offer. A lot of rural America is starting to see some decent speeds offered by WISPs with newer radios. Every year some pockets of of rural America are getting fiber and gigabit speeds. Where might the satellites fall into that mix? We can’t forget that the need for broadband is still doubling every three years, and one has to consider the speeds that homes will want a decade from now – not the speeds households want today. We’re at least a few years from seeing any low-orbit broadband connections and many years away from seeing the swarm of over 10,000 satellites that are planned for broadband delivery.

Modernizing CPNI Rules

I think we badly need new CPNI rules for the industry. CPNI stands for ‘Customer Proprietary Network Information’ and are rules to govern the use of data that telcos and ISPs gather on their customers. CPNI rules are regulated by the FCC and I think it’s fully within their current mandate to update the rules to fit the modern world.

While CPNI is related to privacy issues it’s not exactly the same. CPNI rules involve how ISPs use the customer data that they must gather in order to make the network operate. Originally CPNI rules involved telephone call details – who we called, who called us, etc. Telcos have been prohibited by CPNI rules from using this kind of data without the express consent of a consumer (or else in response to a valid subpoena from law enforcement).

Today the telcos and ISPs gather a lot more information about us than just telephone calling information. For instance, a cellular company not only knows all of your call details, but they know where you are whenever you call, text or make a data connection from your cellphone. Every ISP knows every web search you make since they are the ones routing those requests to the Internet. If you buy newer ISP products like home automation they know all sorts of details that they can gather from monitoring motion detectors and other devices that are part of their service.

Such CPNI data is valuable because it can be used by the ISP to assemble a profile of each customer, particularly when CPNI data is matched with data gathered from other sources. Every large ISP has purchased a business arm that is aimed to help them monetize customer data. The ISPs are all envious of the huge advertising revenues generated by Facebook and Google and want to climb into the advertising game.

The FCC was given the authority to limit how carriers use customer proprietary data, granted by Section 222(b) of the Telecommunications Act of 1934. Those statutes specifically prohibit carriers from using CPNI data for marketing purposes. Over the years the FCC developed more specific CPNI rules that governed telcos. However, the FCC has not updated the specific CPNI rules to cover the wide range of data that ISPs gather on us today. Telcos still ask customers for permission to use their telephone records, but they are not required to get customer permission to track web sites we visit or our location when using a cellphone.

The FCC could invoke CPNI protections for companies that they regulate. It gets dicier for the FCC to expand CPNI rules past traditional carriers. All sorts of web companies also gather information on users. Google makes most of their money through their search engine. They not only charge companies to get higher ranking for Google searches, but they monetize customer data by building profiles of each user that they can market to advertisers. These profiles are supposedly very specific – they can direct advertisers to users who have searched for any specific topic, be it people searching for information about diabetes or those looking to buy a new truck.

There are many who argue that companies like Google should be brought under the same umbrella of rules as ISPs. The ISPs rightfully claim that companies like Google have a major market advantage. But the ISPs clearly prefer the regulatory world where no company is subject to CPNI rules.

There other web applications that are harder to justify as being related to CPNI. For example, a social network like Facebook gathers huge amounts of private data about its users – but those users voluntarily build profiles and share that data freely.

There are more complicated cases such as Amazon, which has been accused of using customer shopping data to develop its own product lines to directly compete with vendors selling on the Amazon platform. The company clearly uses customer data for their own marketing purposes – but Amazon is clearly not a carrier and it would be a huge stretch to pull them under the CPNI rules.

It’s likely that platforms like Facebook or Amazon would have to be regulated with new privacy rules rather than with CPNI rules. That requires an act of Congress, and it’s likely that any new privacy rules would apply to a whole large range of companies that use the web – the approach taken by the European Union.

Broadband and Gaming

I recently saw an interesting statistic that showed that the most popular worldwide video content is gaming. The worldwide gaming video content industry has more than 665 million viewers which makes it bigger than HBO, Netflix, ESPN and Hulu combined. This is a segment of the video industry that I was only peripherally aware of, which I suspect is true for many of you as well.

The GVC (Gaming Video Content) industry is distinct from the on-line playing of games. The GVC content consists of watching others play games along with content that talks about gaming. The industry is estimated to generate $4.6 billion in revenues in 2017. One third of that will come from subscriptions to GVC content along with other direct consumer spending. The rest comes from advertising. There is a whole industry that has sprung up around the GVC content including big conventions and merchandise.

While you can’t characterize such a large group of people, the gamers and GVC viewers are often what you might think of as tech-savvy. In the US the average GVC viewer is around 30, has more education that average and makes a higher than average income of around $58,000. And while you might expect the viewers of GVC content to be largely male a surprising 46% of GVC viewers are female.

Around the world there are numerous video platforms that have been created for gaming content. In the US and Europe the biggest content provider is Twitch. This is a platform that was originally known as Justin.tv. The platform was created in 2007 by Justin Kan and Emmett Shear. The platform allowed users to post live video streams that could be watched by anybody else on the platform. The platform was often used to show pirated live sports feeds, but over time the majority of the content centered around gaming.

Justin.tv was a large content generator and in 2013 – before Netflix really took off – the service said it had 45 million unique viewers and was the fourth largest source of peak Internet traffic in the US. When the biggest competitor to Justin.tv shut down the platform had a near monopoly on gaming content.

The company was renamed to Twitch Interactive and was acquired at the end of 2014 by Amazon. Amazon beefed up the underlying delivery platform, which increased the quality of the streams. Since then Twitch has grown significantly. Amazon reports that the service has over 100 million unique viewers per month, nearly 10 million per day. The average number of simultaneous viewers at any given time on the platform is about 622,000.

Amazon has grown the service by opening up the platform to ‘partners’ much as it has done with OTT content. Twitch now has over 17,000 partners – those that stream unique content. 35% of the content is viewed on cellphones, with the rest on landline broadband connections.

Twitch viewers are loyal. Over half watch the service more than 20 hours per week – and for many of them this is their primary source of video content. The average Twitch user watches the service for 1 hour 46 minutes per day.

While the Twitch platform is free (and I recommend taking a few minutes to watch the above link), many of the channel partners charge monthly subscriptions.

I find it interesting that Twitch is not counted in the universe of OTT providers. But Twitch viewers and statistics are separate from, and not counted with viewers of Amazon Prime. Perhaps this is not considered as OTT content since a lot of the content is viewer-generated. But this is still largely true for YouTube, which is now counted among the OTT providers. Many of the channels on Twitch are now professionally produced and certainly are hard to distinguish from other OTT content.

The GVC industry is worth noting because they are a big source of video content on our broadband networks. The video watched on the web doesn’t just come from sources like Netflix and more and more of it is coming from platforms like Twitch that carry a mountain of viewer-generated content. This is just one more example of how the major programmers are going to be in trouble as the generations turn. Younger viewers are not watching traditional programming to nearly the degree of older generations.

Content Finally is King

One of the more common memes in our industry is the phrase “content is king.” This was first said by Sumner Redstone of Viacom in 1994 but made more famous by Bill Gates in 1996. The phrase has been used since then to describe how the creators of content have the power in our industry – be that programming or web content.

John Stankey, the CEO of AT&T Entertainment, recently emphasized this same concept in talking about the company’s planned merger with Time Warner. At the recent Mobile World Congress in Barcelona he said, “We just cannot envision a future where AT&T is relevant if we don’t directly participate in some of the water flowing through our pipes.”

All of the big ISPs have decided that content is key to their survival. Comcast already owns a mountain of programming, and after the merger with Time Warner, AT&T will be a content powerhouse as well. Verizon has climbed into the game with the acquisitions of AOL and Yahoo. There are web companies with the same philosophy. Netflix has built a new industry by creating new content. Google is pushing content heavily through YouTube. Amazon has started to create unique content and recently said they are going to make that a priority. Facebook is becoming a content force through Facebook Now.

I remember having this conversation with Derrel Duplechin of CCG back in 2000. We were asked by several clients to speculate about the future of the carrier industry and we foresaw that most carriers were likely on the path to eventually become what we called “dumb pipe” providers. I remember that this was a story that many of our clients did not want to hear.

We lived in a different carrier world in 2000. Most homes still had telephones and voice was the most profitable product for most carriers. The cable TV product that many of our clients sold then also had decent margins. But we predicted that both products would eventually sink in importance and in margins and that eventually most of our clients would earn most of their profits from broadband. We thought this would happen to all carriers, small and large, and we figured that the most profitable future companies would be those that found some other line of business other than just selling data pipes to end users.

We had some clients take this to heart and some of them have made a really good living by providing extra value to customers. For example, we have several clients who thrive by bringing a suite of products to businesses other than just plain connectivity. But for the most part, the majority of the ISP industry sells dumb pipes today. They compete with the speed of those pipes and with price and with good customer service – but the primary products (and the driver of most of the profits) are now data pipes.

The big companies like AT&T, Verizon and Comcast looked at that future and it scared them. It’s pretty obvious that if your only product is dumb pipes that your earnings are not going to continue to grow fast enough to satisfy Wall Street. This is probably what convinced Verizon to stop expanding their FiOS network. Both AT&T and Verizon got huge earnings boosts from expanding their cellular businesses, but that industry also seems to be heading towards the same plateau as landline ISPs – cell service is becoming a commodity.

So these big companies are now pursuing content because it looks to be the last area in our industry with the potential for significant bottom line growth. It’s going to be an interesting race to watch. Content providers have succeeded or failed over the years according to their ability to find smash hits. A huge hit movie or TV series can mean huge returns to the bottom line. But content providers that don’t create what the public wants to watch suffer badly in terms of stock prices and earnings. Being a content provider is not predictable in the same way as telecom.

Interestingly. AT&T, Verizon, and Comcast are now direct competitors of Facebook, Google, Amazon and Netflix. Content certainly is king, but content also brings the risk from competition. The companies that fall behind in this race are likely to be gobbled up by their more successful competitors. I find it extremely unlikely that all of these big companies will still be in existence in 10 years.

There is no real barrier to entry into the world of content creation other than having a pile of money. It’s likely that other big companies will join the content fray. But all of these companies are entering a world that is in big flux. For example, traditional video and web content might well be replaced by virtual and enhanced reality. The companies that succeed in content will have to spend a lot of money staying one step ahead of the competition, and my money is on the more nimble technology companies. Twenty years ago I would have been shocked to know that someday AT&T would have a CEO of Entertainment – and that may turn out to be the most important job in the corporation.

Amazon as an ISP?

Amazon EchoI mentioned in a blog last week that there is a rumor that Amazon is considering becoming an ISP. This information came from The Information, which says it got this by somebody inside Amazon management.

It’s an intriguing idea. Amazon has shown throughout its history that it loves to own its supply chain. If you recall, Amazon started out as a web reseller of books. But over time the company has built what must certainly be the largest and most efficient bricks and mortar fulfillment infrastructure in the world.

And the company hasn’t stopped there. The company has been building a fleet of semi-trailers used to haul its inventory, thus bypassing UPS and the Post Office. The company uses third-party tractors today but their goal is to build a fleet in anticipation of self-driving trucks within the coming decade. The company is also experimenting with drones, wheeled robots and other ways to bypass local delivery services.

The company has done the same with its successful data center business. They have built massive data centers and assembled a dark-fiber network to connect them together and to connect to major customers. And it is that fiber network that could create the backbone of an ISP network.

You have to think that Amazon learned a lesson from Google Fiber’s foray into FTTP, and so it seems unlikely that they would leap into a massive infrastructure build in that same mold. The article says Amazon might consider using the open access networks in Europe as a way to avoid building fiber. But they don’t have to go the whole way to Europe to try this. For example, just across the mountains from Seattle are a number of Public Utility Districts (county-wide municipal electric companies) that have built open access fiber networks that pass over 100,000 homes – an easy way for Amazon to test the ISP idea.

And around the country are a number of other open access networks. All of the municipal networks in states like Colorado, Utah and Virginia are required by law to be open access. We have the example of Huntsville, AL that built a FTTP network for Google that will become open access after a few years. There are numerous communities around the country that would gladly build fiber networks if they were guaranteed to get companies like Amazon and Google as major ISP tenants. It’s been my experience that almost no city wants to be an ISP unless it has no other option – but there are many who want fiber badly and would welcome Amazon with open arms.

I would think that Amazon will also keep their eye on the developments with wireless last mile. There might come a time when they might be able to leap into the ISP business with a reasonable cost per customer – at least in selected markets.

Amazon would be an interesting ISP. It was just a few years ago that it was clear that an ISP needed a traditional cable TV product to be successful. Google tried to launch without cable TV in Kansas City and hit a brick wall in selling to residential customers. But the tide is turning and I’m not sure that TV is mandatory any longer.

Amazon already has an impressive content platform with Amazon Prime and they have said that they are going to spend billions to create their own content, following the lead of Netflix. It’s also becoming clear that customers are becoming willing to accept an abbreviated line-up of popular cable channels like what’s being sold by Sling TV and other OTT providers. Amazon could be competitive with an abbreviated cable line-up made up of local programming, popular cable channels and its own content.

But Amazon has some advantages that other ISPs don’t have. For now Amazon is leading the pack in the intelligent personal assistant market with its Amazon Echo. I’ve had an Echo for about six months and I already can tell that it is improving. The company is working towards introducing cloud-based AI to the platform and within a few years the Alexa assistant should become a true computer assistant like has been envisioned for decades in science fiction.

My gut tells me that bundles which focus on smart computer services like Alexa will soon be more popular than the traditional triple-play bundles from Comcast and AT&T. Amazon has one huge advantage as a start-up ISP in that customers like using them – something they have fostered by delivering packages regularly on time to a huge percentage of households in the country. They are at the opposite end of the customer service scale from Comcast and the other big ISPs.

I have no idea if this rumor is true. But the idea is so intriguing that I hope Amazon is considering it. One of the major complaints about broadband in this country is the lack of competition and choice. Companies like Amazon can bring fresh competitive bundles that break away from the traditional triple play and that can redefine the ISP of the future.

Update: This rumor persisted and in February 2017 I posted an update about this rumor. https://potsandpansbyccg.com/2017/02/21/amazon-as-an-isp-2/