I see this a bit differently, and I frequently use the history of Google Fiber as evidence for strong net neutrality regulations.
In the areas where it was deployed, Google Fiber was wildly successful. And yet, Google threw in the towel. Why?
It _isn’t_ because Google Fiber is was unprofitable. To the contrary, it was profitable _despite_ driving broadband prices way down in its service areas.
It’s because the incumbent ISPs simply have too much power. That power comes in many forms:
1. Excessive industry lobbying power;
2. Anti-competitive exclusivity agreements;
3. The ability to stall rollouts indefinitely in the absence of “one touch make ready” policies;
4. Required cooperation with and permission from governments who feel less pressured to move quickly once their residents have at least one incumbent provider;
etc.
If the incumbents are so well protected that one of the most powerful corporations in America found the barriers to entry too burdensome, then how can anyone argue that there is any semblance of a “free market”? What hope is there for startup ISPs?
The incumbents have received hundreds of billions of taxpayer dollars, so they already had a government-assisted head start. And, despite prices being so inflated that they halved overnight wherever Google Fiber rolled out, they are still able to effectively stave off meaningful competition.
Would that be possible in a healthy, competitive marketplace? No. So we need to stop treating it like one, when it clearly isn’t.
Several of your factual assumptions are just wrong. For example, there are no "exclusivity agreements"--those have been illegal under federal law since 1994. Also, ISPs never received "hundreds of billions of taxpayer dollars." That number is a total fabrication: https://news.ycombinator.com/item?id=7709556.
But leaving the facts aside, this is something you can logic through. Instead of speculating about what problems Google might have faced, look to its actual conduct to see what problems it tried to avoid. If "industry lobbying power" was the problem, you'd expect Google would try to deploy fiber somewhere it had the leg-up lobbying-wise, like Seattle or San Francisco or Mountain View. It never did that. Under your theory, you have to assume that Comcast (based in Philadelphia) has more lobbying muscle in Google's own back yard than Google.
What did Google do instead? It went to medium-sized cities in Red states with overhead utility lines. Atlanta, Austin, Provo, Kansas City, etc. Why? Because those cities waived regulatory requirements they applied to incumbents, including the crucial one of build-out requirements (the requirement to build out to the whole city). Google went places like Kansas City, which fast tracked permitting, gave free leases of public land for fiber huts, and waived the obligation to cover low income neighborhoods. Indeed, the willingness to waive build out requirements (which are routine in the industry), is one thing these disparate Fiber cities have in common.
All of this tells you that the basic problem with Fiber was profitability (not just making a profit, but enough profit compared to what else Google could be doing with that money). Build-out requirements kill profitability, because you’re forced to spend a lot of money building to neighborhoods where you can expect few customers. (The major cost of a fiber build is wiring the neighborhood, not hooking up each house.) That’s why a cornerstone of Fiber was the idea of “fiberhoods”—neighborhoods with demonstrated demand for fiber. (Of course, remember, the initial fiberhood selection exluded most low income neighborhoods in KC, as you’d expect. Google was forced to back peddle and add a bunch of low income neighborhoods in. But that misstep tells you a lot about the money math that was driving Fiber.)
> For example, there are no "exclusivity agreements"--those have been illegal under federal law since 1994.
I’m not sure we’re talking about the same kinds of agreements; I’m not referring to agreements between providers and municipalities, but between providers and developers, large residential management companies, and HOA-type entities. I do not know whether those are illegal, but in my experience they are prevalent enough that it doesn’t seem to matter. They still get signed, and presumably at least one party believes such covenants are enforceable (I have no idea whether they are or not).
> If "industry lobbying power" was the problem, you'd expect Google would try to deploy fiber somewhere it had the leg-up lobbying-wise, like Seattle or San Francisco. It never did that. Under your theory, you have to assume that Comcast (based in Philadelphia) has more lobbying muscle in Google's own back yard than Google.
I would expect cities where incumbent ISPs’ political influence is weakest to already be benefiting from increased competition, and indeed, that seems to be the case.
I wouldn’t expect Google to prioritize cities that are already well served by multiple providers.
Landlord exclusivity agreements are not illegal, but they’re also only really an issue for large multi-dwelling units. Only 17% of housing units nationwide are in a building with more than four units (and 70% of housing units are single family homes or mobile homes). It’s just not a driving force outside a handful of the largest cities.
You can also add a layer of conjecture, that there are cities that have more competition because lobbying power is weaker. What makes you think Kansas City or Provo or Austin or Atlanta fit into this category more so than Mountain View? And it’s not like most of Silicon Valley is “already well served by multiple providers”—why not build there? I think it’s obvious that the answer to that is not “lobbying” but “blue state cities make it an expensive pain in the ass to build anything including fiber.”
> but they’re also only really an issue for large multi-dwelling units
Are they? I lived in a large single-family home neighborhood in Florida for several years where the HOA has an exclusivity agreement with one of the big cable companies. Why wouldn't those be an issue as well?
My understanding is you can get the HOA to change its rules through assembling a large enough voting block to do so. It's not like the HOA rules are designed to benefit anyone other than the homeowners which comprise the association or anything.
My sister used to live in a HOA neighborhood and has some stories to tell about power grabs, corruption and lies almost on a Game of Thrones level involving the neighbors vying for control of their own little fiefdom.
>My sister used to live in a HOA neighborhood and has some stories to tell about power grabs, corruption and lies almost on a Game of Thrones level involving the neighbors vying for control of their own little fiefdom.
> "Only 17% of housing units nationwide are in a building with more than four units (and 70% of housing units are single family homes or mobile homes)."
Yes. But ultimately isn't it the number of ppl being served, not the building count, that matters more?
It seems like you are misinterpreting the term housing unit. It isn't the number of buildings. A single apartment in a 200 unit building counts as a housing unit. What this number is telling you is just less than 1 in 5 households in the US live in a building with more than 4 units. I would suspect that the distribution of household size if anything is a bit biased towards single family homes having more people per unit in them if anything.
Yes. But if I'm not mistaken the unit of measure was internet access. Yes, 1 in X might not be a single family house, but if there were enough units under a single 1 then in theory that could be more total ppl / accnts than single fam home.
Or have I somewhat forgotten what the OP + resulting thread was trying to unpack? Sorry?
My large apartment building, in downtown Palo Alto, has 2 sets of wiring: one exclusively for the phone company, and one exclusively for the cable company.
Not surprisingly, my 2 choices for fast Internet are the phone company and the cable company.
At one point the FCC was going to issue a rule opening these wires up. My building would be wired with a half-dozen competing ISPs about 5 seconds after that rule was issued.
Would all of those service providers be rewiring the building at the same time? Wouldn't it actually take months to install all the additional cabling needed? What happens when a 6th, 7th or 8th provider pops up. That's just one apartment building.
What happens when xyz start up wants to dig up a whole neighborhood to put their data cables in. Then a month later a new startup that promises to give you a better experience comes in and rips up the road. What happens when xyz start up fails and the half finished project just sits there? Do the tax payers pay to fix it?
Typically speaking, all of the wiring is done in some central location for the building. The phone/electric/cable companies don't want to make individual drops to each unit. They make one drop to the building and then fan out to the individual units. All that changes when a new provider comes in is that they make a drop to the building and then connect the units wiring to it. They do not need to wire to each individual unit.
The level of complication for making the initial drop to the building will vary depending on the area and how wiring is done in general there, areas where there are pole hung wires making the drops are pretty quick to add areas where the wires are underground are typically a little more complicated.
Not sure if you're joking but with the state of housing in the Bay Area, moving for a reason such as "Slow/expensive internet" isn't really an option.
I'm having an issue with second hand smoke from an apartment getting into my unit and I have to put up with this disgusting problem since moving involves other compromises I'm not willing to make at the moment.
> For example, there are no "exclusivity agreements"--those have been illegal under federal law since 1994.
As you know, there's a big difference between "this is illegal" and "this doesn't happen, because the laws prohibiting it are enforced effectively". Unfortunately, exclusivity agreements do not belong to the latter category. The New York Times, for example, has provided fairly extensive coverage over the last five years of how exclusivity agreements (despite being illegal in New York) have eliminated the possibility of choice in broadband ISPs for many New Yorkers, despite the fact that they may live in buildings that are otherwise already wired for alternative ISPs.
That article is talking about arrangements where the government itself agrees not to compete with the ISP (which may be legal). Exclusivity agreements—where the government makes the ISP the exclusive provider for an area by banning other private entities from competing with the ISP—are illegal. (The ban on exclusivity agreements was part of canle deregulation which focused on creating private competition. Whatever the merits of municipal fiber, it’s a misnomer to say that a government provider adds “competition” to a market.)
If buildout is the big problem running fiber, why don't suburbs see more fiber deployments? Aren't most of the wealthy neighborhoods in a lot of major metros outside the city limits?
Verizon, for example, pretty clearly prioritized suburbs when deploying FiOS across its footprint. E.g. Baltimore doesn’t have it, but almost all of the rest of Maryland does. Suburbs, of course, suffer on the density front. The cheapest way to deploy is the way Google did it (or at least tried to): find medium density cities with overhead utility lines, and deploy first to neighborhoods with demonstrated demand.
Right, so I generally buy this narrative, but I'm left wondering: aren't the wealthy suburbs of Chicago --- forget Chicago itself! --- a comparatively more lucrative market than Charlotte or Nashville? Why would Google wire up Nashville first?
Only if it's cheaper to build in Chicago suburbs than Nashville suburbs. There isn't a lot of pricing power for consumer Internet with alternatives available so Google fiber can't charge twice as much even if it costs twice as much to build. The majority will just crawl back to the cable company if the price is too high (despite their incomes).
In other words, wealthy people aren't really relevant. Density of middle class and upper middle class is what makes these build-outs successful.
Right, I get that the target is "density of households above a threshold income"; I'm just saying I'd be surprised if there were more of those in Nashville (the city) than in the Chicago suburbs. There are 6.8 million people in the Chicago combined MSA excluding Chicago, which is almost exactly 10 times the entire population of Nashville.
Sorry, I wasn't being clear enough. I'm sure the wealth density is there, I'm just suggesting that the regulatory costs of building in Chicago suburbs are higher enough that it made more sense to start in Nashville.
Cynicism: the wheels require a lot more grease in Chicago than they do in Nashville.
Just a note to say that this is an awkward construction. The OMB defines Metropolitan Statistical Areas and Combined Statistical Areas, but nothing called Combined MSA, which seems to be a conflation of the two.
My comment references the original source where that number came from, and criticises the methodology. You can read it yourself and see what the number really represents.
But notably, Provo is very red. They also had an existing fiber network infrastructure that they basically gave to Google in exchange for the promise of offering a free service tier to residents.
Same for Atlanta. IIRC the few areas in Atlanta that ended up getting Google Fiber are also African American (the historic MLK area) —- interestingly gentrification has kicked into extreme.
Not only are the incumbents well protected, but they are increasingly few.
2 years ago I had 300 Mbps cable service at home via Time Warner Cable, then they were bought by Spectrum. Now I have 150 Mbps service.
I still live in the same place and I'm still using the same cable modem (one I own, so as to avoid the ridiculous rental fees) I used back then, but 300 Mbps was simply killed as an option. I can't even pay more for it.
Google Fiber may have been a useful gambit to increase the pace of rollout of other telecom's high speed offerings, but I would bet good money that it will be a very short-term play with erosion of net neutrality and further consolidation of the telecom market quickly wiping out the benefits.
Technical POV: 300Mbps might have been possible two years ago because there were fewer high bandwidth customers on your segment of the DOCSIS3 network. With increased channel bonding, it's now possible to do gigabit downstream over cable. For an example, Google "24x8 docsis3 modem". You may have something like an 8x4 modem. It is an intentionally oversubscribed shared access medium on the copper you're connected to, so it will only support a finite number of 300Mbps customers. It will support more 150Mbps customers on the same segment.
The upstream still has to be highly asymmetrical because they're allocating rf channels asymmetrically. The capacity of one segment of coax is still a great deal less than a GPON network, nevermind active Ethernet.
In other countries, speeds continue to increase, even as the user base grows over the years. So this is kind of a BS argument. ISPs should be investing more into their infrastructures. Instead they use the money for buyouts and to pay shareholder dividends.
You'd think by now they would finally start having a bit more upstream traffic so they could even out the assymetry a bit. But here I am on my 500mbps down, 30mbps up connection.
I'm curious did Spectrum send you a letter after the acquisition explaining that your service was being halved? Did the price go down with it or are you just paying more for less? This is the first I have heard of this.
>> If the incumbents are so well protected that one of the most powerful corporations in America found the barriers to entry too burdensome, then how can anyone argue that there is any semblance of a “free market”?
Yes, that's the biggest problem and I'm tired of repeating it. If Google can't do it, then what hope do the other minnows and plebeians have? Its high time that the average lazy American should wake up and revolt against the establishment for this injustice now.
People need to 1) have a general interest in many areas and 2) Hold their politicions accountable.
Currently it almost feels as though corruption is becoming acceptable. I keep seeing people say:
"they're all corrupt so its political to arrest them for the less serious things."
This seems a sign of the system breaking.
Also without trying to kick of tribalism here, I think wealth inequality drives this corrupt and short term behaviour somewhat. People start relaxing ethics as their focus turns to getting ahead. I feel if everyone is more balanced economically other things become more important for your feelings of self importance such as being judged for the person you are rather than your net worth.
> Google threw in the towel. Why? [...] It’s because the incumbent ISPs simply have too much power
Is this why? I am not aware of any disclosed reasons. I was under the impression that they ended their experiment and realized being an ISP wasn't in line with their business model, profits weren't large enough, and the execs just didn't want to be in that market. Sure incumbency and lobbying happened (and won't change with net neutrality and happens in any entrenched market they'd enter), but is that why Google stopped rollouts? Do you have evidence of this? And if not, why is your speculation more accurate than mine? Also, if this is not for sure why they stopped, do you believe other efforts to promote net neutrality are harmed by frequently using examples like this that may be inaccurate to assist?
I think that free competition alone is responsible for having cheaper and better internet access: in Romania I pay 10$/month for a 1gb, unlimited traffic, internet connection. I think we have this for about 3-5 years.
In a country with very few companies that are bigger than 1% of what Google or AOL or AT&T are.
In a country where an entrepreneur doesn't even dream of having access to the big investor money needed to start an optical fiber ISP.
What you're paying in US is 10x-100x more expensive than that. One can only imagine the margins your ISPs make and the profit opportunities other entrepreneurs have in this space.
If it's not about the incumbent's overwhelming power, what can it be? Lack of entrepreneurial spirit in US? Lack of technological knowledge?(you invented this whole internet thing ffs) Lack of desire for easy profits?
> If it's not about the incumbent's overwhelming power, what can it be?
Building fiber is really expensive in developed countries. Look at it this way. According to Google, a round-trip ticket on the Bucharest Metro is about $1.25. On the DC Metro, it’s $4-12 depending on time of day/distance. ($5 each way during commuting hours from the suburbs is typical). The DC Metro must be hugely profitable, right? Well no, actually it loses a lot of money!
The key metric for telecoms is ARPU - average revenue per user. Net profit is a pointless number. Ebitda margins are something like 36% for verizon from a recent google search.
From my time making telecom comparables - you want to be America. American ARPU numbers aresignificantly higher that the world average. America had something like $40 in 2006, to $35 in 2016.
IN contrast, Germany has 14$ today, and Asia something around a $1 - $3.
Dont have access to any industry finance/research papers.
America telecom is moribund, To argue otherwise is to mislead. Google failed to change the market. Despite dumping huge amounts of money into it, something no other firm can afford to do.
Where it entered, incumbents magically were able to offer better services which they had avoided providing.
THis HBR report reads like MBA speak, more a lesson on spin than actual fact.
Remember when these firms hid zip code information in auctions so that they wouldnt cross bid?
I read the history of American telecom with bemusement, the lessons highlighting the necessity to build good regulatory frameworks so that the economy can work efficiently.
Finally - Building fibre is difficult, even in developing countries. The fact that you have regulation stopping people from tearing up the road, or just having wires crossing over houses willy nilly is the happy cost of being a civilized nation.
> The key metric for telecoms is ARPU - average revenue per user. Net profit is a pointless number. Ebitda margins are something like 36% for verizon from a recent google search.
OP’s assertion was that because fiber service is much more expensive in the US than in Romania, American telecoms must be making huge profits on it. ARPU—which ignores costs—is useless for analyzing that assertion.
EBITDA margin is also useless for looking at telecom profitability because it’s exclusive of depreciation and interest. Telcos spent tens of billions of dollars on fiber that’s sitting in the ground depreciating. EBITDA ignores that. If anything, looking at EBITDA shows you how dismal the economics of wireline really are. People have this idea that wireline networks are free money after paying for the initial build. In reality, 2/3 of every dollar you pay in service charges goes to ongoing expenses, even before you start accounting for the cost of building the network in thr first place.
> in Romania I pay 10$/month for a 1gb, unlimited traffic, internet connection. I think we have this for about 3-5 years.
1GB is fairly new in Romania, but already by 2007 fiber had been rolled out in the major cities (wise use of EU funds?) and at least 300Mbps or 500Mbps connections for 10€/month have long been routine.
(Incidentally, even though gigabit is becoming common, not many people realize they have to upgrade their routers – wifi in cafes is almost never gigabit, even if the ISP’s link supports it, just because the cafe owners will simply grab a cheap router from the shop and that is unlikely to be gigabit.)
Even before wide availability of fiber, cable and DSL connections offered pretty high speeds for that medium. In Cluj your ISP would even give you the username and password for a private server where you could share films and music with other people in the same city, so of course download speeds would be extremely fast. Good times.
You're making the same mistake as GP by guessing which harms the argument in favor of NN. There are many reasons not to be in a market. The profits are not "easy". Providers charge what they can in markets, so not every dollar cost, especially when used in comparison with a completely different country, is a result of competition. For example, many European countries likely have cheaper and better mobile coverage than the US compared as a whole. By your invalid logic, higher prices and substandard quality in the US is due solely to lack of competition which is false. Same with many other markets and industries you can compare.
These kinds of different-environment correlation mistakes based on some misplaced common sense do not help. More specifically to the Google case, these assumption mistakes that are then used to buttress NN arguments reduce the strength of the argument, especially while there are many evidence based examples in favor of NN one could use instead.
How is NN supposed to fix the entry barrier issues? I think you need a very different sort of regulations (anti-monopoly) and deregulations (barrier removal).
I don't want to argue whenever NN is good or not so much (I'm not sure myself), but my point is the effort seems massively misdirected. Everyone's raving about not policing traffic; states are passing regulations, but (across the ocean, I admit) I don't see anything happening so you guys can get closer to having lots of new ISPs and actual competition.
As I recall, American telecom firms have some of the highest revenue per user - this is the key metric that all telcos get compared on (apples to apples metric). American telcos have the highest ARPU in the world. If they are not profitable, then the rest of the world is positively broke -
American telecoms are not a comparison. Google had to pay for all the infrastructure to roll out the fiber, whereas others were susbsidized by the government.
Telcos are compared to each other, the same way car firms are - that’s the essence of a comp.
Google fiber is, in the end, compared to the same companies because they had to make the same capex. The subsidy will be accounted for in the financial calculus.
As a libertarian who argues for government regulation being the problem and not the market a lot, I must say, I'm not surprised.
This is basically normal and what you expect. People are just blind to the massive amount of regulation that struggles and often basically predetermines the market and when you try to explain how 3 different regulation combine to make the improvement impossible peoples eyes gloss over.
If the regulation is not "THIS GUY HAS A MONOPOLY" its very hard to get anybody exited or even talking about these regulations.
This is not an uncommon argument. It's not that the free market never leads to monopolies - but almost all of the most recognizable monopolies either owe their business model to a distortion in the market directly caused by regulation, owe their continued long-term existence to a regulatory environment that makes it unusually hard for would-be competitors to even think about trying to fight them, or owe their particular reach and dominance to special advantages imposed from outside.
Monopolies, oligopolies, and cartels can (and would) certainly occur even in the mythical Ancapistan, but accidental and deliberate abuses of state power make them both more common, more stable, and worse than they would be otherwise.
Historically monopoly’s show up in many market which seems like a strong counter argument. Companies like Intel have intentionally avoided becoming a full monopoly due to regulations, but economies of scale end up being a huge advantage.
Walmart needs no help to kill off competition in a small town as a more local example.
But notice how Walmart's "monopoly" is different from Comcast's. Comcast has a monopoly because barriers are high. Walmart has one because they're the fiercest competitor. If Walmart raised prices, competing stores would appear immediately. When Comcast raises prices, everyone else has to tighten their belt.
Companies can’t start up and fail for zero cost. Which allows Walmart to raise prices without instantly getting competition in those areas. In higher population areas that can support more stores it’s a very different story.
> Companies can’t start up and fail for zero cost.
Retailing companies? Sure they can. If Walmart started selling products for $5 that retail for $1 elsewhere, anyone could sell the ten they already have in their kitchen to their neighbors for $4.50 each, then use the money they just made to drive to the next town and buy twice as many. Before they've even run out of vacation days at their real job they're doing half the business in the town and can afford to rent a storefront month to month and hire some college students to mind the store.
The second Walmart lowers their prices again, they sell their remaining inventory for the normal retail price instead of the inflated one and then shut down. Assuming they haven't built enough volume and loyalty in the local market by then to continue operating and keep half the town's business indefinitely.
In a ~3% margin business the difference between 1$ and 1.05$ is huge. But, nobody is going to be selling canned tuna out of their basement for under 10 cents profit on the can.
Which is the real power of monopoly pricing, it's not obvious that spending 1.05$ is vastly worse than 1$ but it makes a huge difference to a companies bottom line.
> In a ~3% margin business the difference between 1$ and 1.05$ is huge. But, nobody is going to be selling canned tuna out of their basement for under 10 cents profit on the can.
The margins are only ~3% because the barrier to entry is low. Imagine what their margins would be if they had an actual monopoly on food.
> Which is the real power of monopoly pricing, it's not obvious that spending 1.05$ is vastly worse than 1$ but it makes a huge difference to a companies bottom line.
It's not obvious that spending $1.05 is vastly worse than $1 because it isn't. It's not even clear that the hypothetical five cents even exists -- if Walmart is charging $1.05 because their cost is $1, and they have no competitors because the next most efficient retailer has a cost of $1.07, Walmart is charging you less than the competitor(s) they put out of business, not more.
And that would still be true if they reduced their own cost to $.25 without lowering prices further, when no competitor could reduce costs in the same way, though that rarely actually happens. Squeezing 5% more out of a supplier is a lot more realistic than squeezing 75% more out of them.
Walmart prices do vary. In some places grocery competitors like Aldi do shade many of their prices significantly (and consistently, not just loss leaders).
Prices in different stores vary for a lot of reasons. One area will have higher real estate costs or higher taxes, or a lower volume of customers to amortize fixed costs over.
This is also why a new competitor opening up seems to cause prices to decline significantly. Part of it is the competition, but part of it is that the same factors that allowed the competitor to justify opening a new store (e.g. a tax cut or population growth) also allow the existing store to reduce prices.
I think the argument is that these companies became large enough to afford to buy the regulations that allowed them to discourage new entrants to the market. Then they consolidated the remaining players.
perhaps true of "big L" libertarianism, but this is not entirely a fair representation of the broader political philosophy. most libertarians i have talked to will certainly acknowledge that market failures happen even in unregulated markets.
as i understand it, the core motivation for libertarianism is that power is real and impossible to "distribute" evenly. relatively more powerful people will always be able to divert state resources to serve their own ends. the only way to limit this effect is to limit the power of the state itself.
so the argument is not so much that free markets are beautiful, perfect things; rather, the free market is usually preferable to the regulated markets that powerful actors will allow to occur.
That is true for the right-libertarianism dominant in the US (also true for some flavors of left-libertarianism). I guess that's probably what most refer to with "big-L Libertarianism".
Google has long cultivated a substantial lobbying effort in Washington. In fact it outspent any other single private company in it's lobbying in 2017. [1]
>"2. Anti-competitive exclusivity agreements;"
There's no such thing at either the municipal level, state level or otherwise. Exclusivity agreements for carriers don't exist.
>"3. The ability to stall rollouts indefinitely in the absence of “one touch make ready” policies;"
Telephone poles which were an issue in the Austin rollout are not an issue in many(most?) US cities where there are no telephone poles and cable vaults and ducting are underground and not owned by the incumbents.
>"4. Required cooperation with and permission from governments who feel less pressured to move quickly once their residents have at least one incumbent provider;"
Is there any evidence of this? Politician love to take credit for innovations and improvements that they had very little to do with.
As an ISP: right of way and economics of $ per last mile customer reached are HARD. The incumbents in an area, generally the copper phone line company (ILEC) and cable TV operator (Comcast, RCN, Charter, whatever) were handed the aerial ROW and underground ROW on a silver platter 30+ years ago, when there was no such thing as telecom competition. It was just assumed at that time that those two companies would be installing stuff on the same poles as the power company.
They built wherever they pleased, whenever they pleased, at minimal regulatory cost and with minimal additional costs. Because of course nobody would buy a house with no dialtone phone and/or no cable TV.
Now in a modern era if a FTTH ISP wants to come along and overbuild the same routes as the phone company, it's a lot more difficult, time consuming and expensive.
The reasons this is getting off the ground at all is because our city government spent the past 5 years dotting every i and crossing every t to be able to drop the hammer on AT&T and WOW! without retaliation. Also, we own our own electric utility, and on top of their ROW they're also bankrolling most of this operation with a sweetheart loan deal (overpaying for electricity worked out in the end, who knew).
Having the electrical utility controlled by the local county can be a huge advantage. The counties in eastern WA where the local PUD (public utility district) owns the last mile AC power distribution networks have been generally quite successful building their own dark fiber networks. Because of course they don't need to ask themselves for permission to build, they already own the poles and ROW, they just do it. They already have linemen and bucket trucks, flaggers, insurance, spool trailers, etc.
Then ISPs can reach end user customers by paying the PUD for L2 transport across their networks.
This new PUD fiber infrastructure exists separately from whatever the local Telco and cable TV operators have built.
It also has the benefit where citizens can decide their generation supply (ie renewables) vs shareholders in a for-profit electrical utility or generator deciding based solely on profitability.
For those interested, thefoa.org website also has a lot of excellent free training material and reference related to fiber networks (at OSI model layer 1).
> But what if the company’s goal was never to unleash the disrupter itself so much as to encourage incumbent broadband providers to do so, helping Google’s expansion in adjacent markets such as video and emerging markets including smart homes?
Did anyone think anything else WAS the strategy? Google has a history of such "spurring" efforts. Glass, Fiber, Nexus, heck, even a ton of Android support itself.
Yeah I don't understand at all why people think that Google wanted to be in this business. They were forced to. ISPs were starting to lag behind on bandwidth that was important to them, and also started to discriminate on traffic. It is not a failure at all in any way because they never intended on completely taking over. They'd much rather be in their business.
The same is true with Android, as they were under threat of locked out of a smartphone boom by Apple. They have stuck with it because it has become more popular than they could have ever imagined. I also think this is their reason behind making Kubernetes. I think they want to commoditize the cloud more to avoid dominance from Amazon.
The first android designs were more like Blackberry's than an iPhone[1]. The designs ended up being publicized during the Oracle trial. You can clearly see that they were no where near what Apple was doing with the iPhone. After Apple released the iPhone Google decided to not even release that version and decided to re-do the phone[2]. They believed that a touch screen can't completely replace physical buttons[3]. I'm always surprised how reluctant people are to give credit to Apple for completely changing the smartphone industry.
Are you new around here? There’s a vocal set of diehard contrarians that insist LG was the true pioneer and Apple just has better marketing, the iPhone wasn’t a better product.
It’s the most successful consumer electronics product in history, but some still are in denial.
The evolution of the smartphone market was obvious even in 2001/2002, if you saw a Handspring Treo 180 and used one on a GPRS/EDGE network. I had one. When I saw the first info about Android, it seemed to me like a logical technological advancement in the same field. At the time in 2004/2005, the iPhone OS was a total unknown, maybe a tiny bit of info had leaked from Apple about their secret project, WinCE/Windows Mobile was a joke, palmos was a dead end.
I disagree. The iPhone was significant leap: a captive, multi touch screen (not resistive), a full desktop OS (hidden from users, but not a scaled down embedded OS like everyone else), finger, not stylus focused, touch, not button focused.
Yet at the launch I wasn't impressed ("doesn't solve a problem I have"; clearly I lack vision as it took until iPhone 4 for me to see the light).
> Yeah I don't understand at all why people think that Google wanted to be in this business.
People still think Google Glass was a failed product. Many just have no idea how businesses, particularly non-traditional ones like Google with an appetite for change and a pocketbook to buy what it wants, are run.
i mean, if a person misunderstands the goals of any project, they are likely to see its outcome as a failure. that doesn't really say anything about the project, though.
Verizon makes roughly $100 billion a year off of internet subscription revenue, and collects immense amounts of user data.
If Google went through all the investment to build an ISP, bid on spectrum, etc., it'd be silly for Google to at least not cautiously/optimistically aim for the main prize of the market.
It seems more likely that competition was too costly, and it was fine falling back on the modest goal of prodding the market forward.
No, this is incorrect, they spent a ton of money and time building new last mile aerial fiber. Don't confuse long haul dark fiber IRUs with the last mile access layer of the network.
Are you sure about "most"? Google has an extensive backbone network, but they didn't have fiber to every house in every neighborhood they offer service in.
That is definitely what I thought the strategy was. OTOH, a lot of people thought that the Google Chrome strategy was just to spur other browser to advance. In the end, they've gained tremendous market share with the product.
Nexus had a U shaped curve in pricing. Earlier ones were too expensive. Got cheap in the middle, but again became expensive and morphed into the pixel line. It's only in the middle (around 2012-14) where they were meant to attract developers and the masses.
I'm going off of my personal take only, so it's pretty safe to assume I'm wrong, but the effort seemed so half-hearted that I assumed Google was pushing other makers to stop installing bloatware and get on top of Android updates by providing JUST ENOUGH competition to force them to take it up. Until then the major players were all bad about those and profited from it so there was no incentive for them to change their ways and consumers had no real way to talk with their wallets on the topics.
Not that todays' options are GOOD on those points, but they've definitely improved, at least with Android version updates.
Looking at the outcome after the fact, it's very easy to attribute an action to some master plan to achieve that outcome indirectly. I prefer to believe that 99% of the time, when someone says they're launching an ISP, they're doing it to become an ISP. Here's the initial Google Fiber blog post so you can judge for yourself:
i remember reading coverage of google fiber in tech publications when it first rolled out. even then they were describing it more as a "swift kick in the ass" to incumbent ISPs than a serious attempt to corner the market. it's like the early intel consumer ssds. intel never wanted to be a big (ie low margin) ssd vendor; they wanted to stimulate the other players to make faster storage to unbottleneck their processors.
even your linked blog post seems to support this point. they conservatively couch the product as an "experiment".
> We don't think we have all the answers – but through our trial, we hope to make a meaningful contribution to the shared goal of delivering faster and better Internet for everyone.
does that sound like the PR blurb of a company that seriously intends to offer hi-speed internet to every person in america?
What a great bargaining chip to have if one were complaining to various telecom companies about how they well they carry Google data across their networks.
Let's not forget Chrome. Mozilla wrestled about 25% of browser market share away from Microsoft and IE with Firefox but seemed unable to get past that. Chrome came along and now IE is dead and Edge is an afterthought.
> But in 2009, no leading carrier was planning a major upgrade of its existing physical plant.
That’s not true (and the article doesn’t cite to any support for that idea or any other). In 2009 ATT and Verizon were in the middle of deploying Uverse and FiOS, which were planned since 2005. By that time, Qwest (later acquired by CenturyLink) had already passed a million households and had service in a dozen cities.
The article credits Google with these deployments, but makes no effort to show that the pace of these projects, which were already in the pipeline, somehow accelerated after Fiber. This would be an easy assertion to prove: show there was a discontinuity in FTTx deployment after 2010. The article states that there is 30% urban FTTx penetration today, but doesn’t say whether that’s on or off the pre-2010 trend. (The article also gets the cable trajectory wrong. In 2009, everyone was in the middle of deploying DOCSIS 3.0. The pace of DOCSIS upgrades didn’t speed up after Fiber. DOCSIS 3.1, and gigabit cable, came when you’d expect.)
What Google does deserve credit for, I think, is making gigabit happen. ISPs likely would’ve held that as a premium tier for market segmention purposes. After Fiber set the gigabit at $70 price point, everyone had to match (or come close). That was something that wasn’t in the road maps before Fiber.
>ISPs likely would’ve held that as a premium tier for market segmention purposes
They still do. I don't think much has changed regarding actual fiber. Comcast, for example, will sell you passable gigabit over cable in the < $150 category (not cheap), but that has some abysmal upload (probably 20 Mbps), and lots of bufferbloat. If you want real fiber from comcast, be prepared to pay $500 in installation, $300 a month, and a 2 year contract. This is in ATL, which is a google fiber market, just that google fiber never really rolled out beyond the google fiber shop.
Comcast Business quoted me gigabit symmetrical for $1k install and around $500/month with a 3 year contract at my location. Plus equipment costs. Completely unacceptable.
Seattle, BTW. Where CenturyLink is theoretically rolling out gigabit, but not actually.
The few places I’ve spot checked - including areas that I know are considered less affluent in the south metro area are covered by AT&T gigabit up and down for $70-$80 a month. I have AT&T Fiber in the northern burbs and routinely get 959/950.
FiOS was winding down expanding to new markets, but was squarely in the middle of the overall build out. The initial plan was to cover 18 million homes. Of that, about 13 million were complete by 2008. In 2009, Verizon was already in negotiations with New York. After Fiber came out in 2010, Verizon continued the planned build out at a slower pace, ultimately passing 20 million houses. What was the measurable impact of Fiber on that deployment? It probably did spur the gigabit tier, which required more oversubscription than Verizon planned for originally.
> The article credits Google with these deployments, but makes no effort to show that the pace of these projects, which were already in the pipeline, somehow accelerated after Fiber.
Do you live in an area that was targeted for a Google Fiber rollout? I do. Weeks after Google announced our area was getting fiber — BAM! AT&T and TWC dropped prices and raised speeds dramatically . Not long after that AT&T had run FTTH to a significant number of households. Previously that had been unheard of in this area. Google most certainly lit a fire under their asses.
Though Google appears to have paused future deployments, the broadband business
has permanently changed. Fiber investments by former telephone companies have
accelerated or restarted.
As someone who lives in Austin, TX and walk past the Google Fiber office every day, I can testimony that thanks to this mindset now there are at least 3 ISPs serving Fiber to the Home. (Grande, ATT, Google and possibly Spectrum?)
Even if you can't get Fiber, most gentrified blocks can get 300Mbps (and with offers you can get it for $50ish a month). That's amazing, just because ISPs were afraid of Google's offering.
All in all Google's Fiber is super limited (very few blocks can get Fiber to their home) but just like a "grimreaper waiting on your doorstep" ISPs competing with Google had to start offering really good deals to everyone.
I live in Austin. For two years I waited for my neighborhood to be included in their coverage map. Even after the map showed I was covered, repeated attempts to sign up would get rejected by their automated registration system. For months I watched workmen laying in the orange cables, including right through my front yard.
In October 2017 a Google employee walked door to door in the neighborhood signing people up. We signed up, gave a $15 (?) deposit, and then never heard from him or google again.
On the other hand, in that span of time my Spectrum speed got bumped repeated from 15/5 Mbps to 200/12 Mbps, without changing my price. So, thanks Google for bluffing so well.
I lived in Austin and had AT&T and Google Fiber at times. The service, speed, and reliability was exactly the same as was the price. It's amazing what some competition will do. I now live in Minneapolis and actually have 10 Gbps up/down and the option for 1 Gbps up/down for half the price of Google/AT&T in Austin ($35/month), but no Google Fiber (US Internet is the provider here).
I also live in Austin, in the one neighborhood where Grande offers FTTH. Before Google it cost $89/mo for 100Mbps. As soon as Google announced Austin, I could get 1Gbps for $59/mo.
I still wish Google would continue the buildout in Austin, as Grande doesn't have a clue about IPv6.
It is worth mentioning that grande is now owned by the same capital group that owns Wave, which has been a pioneer in affordable gigabit access in condos and apartments, where they can reach a premises economically (flat rate $60/100Mbps, or $80/1000Mbps), no caps, no quotas, no upstream rate limits.
>"The overall goal: to ensure at least 100,000,000 Americans had access to broadband speeds of 100 Mbps by 2020. As it turned out, providers blew past that milestone as early as 2016."
First gripe....why oh why could they not make that goal for a symmetrical 100Mbps and not just focused on the DL speed.
If the push is to 5G, I'd welcome it, if the mobile carriers opened ports on consumer plans. I've cherished my coveted unlimited iPad data plan with att used in my mifi with a yagi that points right at the service tower. It regularly sees over 250Gigs a month (check my twitter for that pic).
Gotta give credit to Google here, regardless what you think about their intentions or competence. They moved the needle in a way that only a disruptor can move needles. (They didn't plan it that way ahead of time, you'll just have to trust me on that.)
Remember they've done this before. Usenet.
And for a successful success story, how about gmail? Just like no one remembers alta vista, lycos, etc (or myspace in more recent history), no one remembers that when gmail launched and offered 1GB of mail, that was absurd. The previous free tier was more like 5MB -- 3 orders of magnitude lower! Then when Yahoo finally came around (had to stop the bleeding) and also offered 1GB, gmail went unlimited!
They are going to write this exact same story in 10 years if electric cars take off (I don't think they will, but leave that aside) and Tesla is dead. Tesla's failure will be the industry's success.
The two-tiered market of high-speed cable and lower-speed DSL broadband has given way to a free-for-all, forcing adoption of more disruptive strategies by incumbents and new entrants alike. The result is increased competition between providers and among cities and regions eager for game-changing private investment.
In which alternate universe is this happening? Certainly not in the United States. How many residential locations have a choice of more than two data providers? Yes, there's wireless, with "Unlimited", "Premium Unlimited", and "Extra Premium Unlimited."
I have gigabit fiber internet (up and down) in LA for $80/m from AT&T. It's incredible, and I can't believe they offer it. I can't imagine that would be the case today if Google Fiber never existed.
From the sounds of it, they had plenty of cities lining up to get Fiber from Google. I would assume that, at least in those cities, they wouldn't face too many regulatory hurdles. So, what's stopping Google from expanding into the willing cities? Is it the expense of digging up all that last mile infrastructure? Is that the limiting factor?
The limiting factor is Google has undergone the same transformation as nearly all market-leading companies go - their function has changed from innovation and disruption to protecting their markets and rent-seeking. They're too large now to do anything without it being prohibitively difficult and expensive.
The day I read an article announcing they'd hired the former CFO of Morgan Stanley, and she would be focusing on "cost containment", I knew Google Fiber was dead.
well, they can always come up with a new estimate. If they'r e not doing it, then it means they can't get a good ROI on it. I'm just wondering what they're biggest expense is: I suspect it's last mile digging up all those trenches - i hear that's really expensive.
One major issue, at least here in Provo, was that their only high-speed offering (at first) was $70/mo. This is a high-proportion college town, and almost no students can afford that (even split amongst themselves). Even if they could, almost nobody wanted to because why pay for that when the free plan works good enough? A speed test at fast.com reveals that Netflix still streams at gigabit speeds even on the free plans. YouTube is similar. (They seem to open the fire hose on those sites!)
Now they have a fifty dollar 100 Mbps plan. But students don't pay for that, either, because they don't know about it.
It was not a well-thought-out business plan.
Their seven-year contract will be expiring soon. Will be interesting to see what happens then.
$50 for 100 Mbps is still bordering on expensive. Comcast undercuts that, at least in the promo period. They jack it up after the period, but that's the usual thing with Comcast. You need to renegotiate or change accounts. So the less informed as well as cheapskates would still go with Comcast. What you get with GF is symmetric 100Mbps and better quality, but hardly anyone would know that.
"But what if the company’s goal was never to unleash the disrupter itself so much as to encourage incumbent broadband providers to do so, helping Google’s expansion in adjacent markets such as video and emerging markets including smart homes?"
I don't think it's a what if. If you can think back to the state of broadband (e.g., Verizon, Comcast, and mobile speeds) when Fiber was announced, and the following the subsequent increases, there's definitely a strong correlation to Fiber's influence.
Goggle got what they wanted by pretending to do something it had no interest in doing. That's not a failure. It's damn clever.
Fiber is great. But... in my view, the key to omnipresent BB is wireless, yes 5G, sorry to grind this ax but, I do believe that is the key, particularly in a country like the USA where there are large geographic areas, it is simply not economic to cover the entire area with fiber. mm-wave is part of the equation but also, lower frequency bands (sub-6Ghz) are key. FCC opening up more spectrum (all bands) is key to wide spread BB access.
Hoping wireless will solve the B.B. problem is laughable. It doesn’t matter what technology you use, the airwaves are a single shared medium while wires are essentially private sanctums of bandwidth within their own little environment with essentially the same bandwidth in each wire. Wireless companies are still charging per Gb, while wireline companies charge per 100s of Gb. People use streaming all the time, and that’s only increasing. Wireless is not going to be viable for the streaming future of 4K, etc for a very long time, and even when the tech is ready, the providers are still going to charge huge sums for that access. If wireless really is cheaper to build out, then why the huge disparity in price per Gb?
One of the keys here that a lot of people miss is that mm-wave alone will not fill the BB need. It is the combination of sub-6ghz and mm-wave. 5Gnr works in high band, mm-wave and low/mid-band. This is the key, using each in the appropriate situation. mm-wave in dense urban, fiber rich areas, sub-6 in more rural and xburbs.
i can see that there is a public benefit in having accessible broadband for every citizen. it is important that people out in the sticks still have access to the news, for example. but is it really of critical importance that everyone have access to 4K netflix streams?
RF guy here doing mnWave circuitry. Don’t count on it. A simple tree will ruin your link, as will rain. 6 GHz doesn’t have near the capacity. One DWDM fiber channel has +100 GHz bandwidth; that’s where it’s at. Leave wireless for what it’s good for; mobility. It sucks for real capacity.
It is still pretty amazing that we now have 10Gbps e-band radios good for 2 to 3 km bridges between rooftops. For $15k can be a lot better than a $200k fiber build. The the total bps/hz capacity of single mode is truly astounding compared to what it's possible to push through the air, FDD style, even with 2000MHz channels.
Definitely, and I think there are even 30 km links at E-band. What concerns me is fiber build-outs are stalling, thinking that wireless is going to save costs. It’s not going to happen, at least not the performance of a fiber to every home. Electricity and twisted pair could be run to every home, but now it’s too costly.
How long will it be before typical households need the performance of fiber? I’ve got 4 gig symmetrical to my house (3 gig Comcast, 1 gig Verizon). I had a wiring issue where my desktop’s ethernet port was syncing at 100 mbps, and I didn’t even notice. The Web can’t even really use the bandwidth—the TCP windowing algorithm takes too long to ramp up. 100 mbps is plenty to stream 4k on the TV plus 1080p to a couple of iPads. Even games download plenty fast at 100 mbps. There is the cloud, but I can back up my entire iPhone to the cloud in less than an hour at 100 mbps—doing it incrementally, it doesn’t even register.
There is a cost to doing fiber build outs before there is a real demand for it. If people don’t need more right now than what cable and wireless can offer, you might wait a very long time to recover your investment.
The cost is in the horizontal drilling, not the actual fiber equipment. Why it could be done “then”, with electricity and twisted copper pair, but not now, is what irks me.
I could use symmetrical 10G right now for software defined radio experimentation. I presently have 100/5, and it’s the 5 that is the issue. Without fiber, we’ll be stuck with horrible asymmetry.
> with electricity and twisted copper pair, but not now, is what irks me
Electricity and telephones brought immediate, game changing benefits, and there was no alternative that was almost as good. Fiber still doesn’t offer “must have” benefits over existing alternatives for typical consumers. Maybe in the future with VR and whatnot, but not yet.
The demand for faster wired connections has stalled even on the LAN. Apple added GigE to the PowerMac in 2000, a year after the standard for 1000base-t was ratified. It’s been more then a decade since 10Gbase-t was ratified, and Apple just added it to a workstation this year. There has been almost no uptake outside the server room.
Lack of demand means that it’s hard to draw consumers away from existing options. FTTH uptake rates are below 50% (meaning less than half of homes who can get it actually subscribe, or put another way you have to build fiber past two or three homes to get one paying subscriber).
On the supply side, anything involving human labor is relatively more expensive compared to other inputs than it was then. Minimum enlisted military pay at the time of the rural electrification act was $9,000 per year (adjusted for CPI inflation). It’s double that today.
Realistically you can't go even 8km reliably in e-band, no way for 30... 30 at traditional frequencies like 11GHz, sure. But the atmospheric attenuation is extreme at 71-86GHz, as is the rain fade. Even with +15-18 tx power and 52db gain antennas.
I swear Facebook claimed 10G at 30 km, and recently Bridgewave was claiming the same. Facebook was using a custom 33 dBm PA MMIC.
Quorvo (or maybe HRL) presented a 33 dBm GaN MMIC a couple years ago, but it was ~90 nm GaN process which isn’t out yet. Once that’s out, I’d expect >1W at 80 GHz to allow some cool stuff.
The hard part is that increased tx power beyond about +18 also increases noise - 33dBm might work for OOK or BPSK modulation with a 5000MHz wide channel each way, FDD, but the commercial 10Gbps radios available now use from 64-128-256QAM to achieve 10GbE in reasonable channel sizes. Using all 5000MHz works, but limits the number of radios you can Colo on one roof. Ever seen the first generation of Bridgewave and gigabeam 71-86GHz radios? OOK modulated for 1GbE.
It's also possible FB's tests are using channels even wider than 5GHz with STA from the FCC.
> The broadband market was experiencing a classic "prisoner's dilemma," where neither cable nor DSL providers felt a competitive threat from the other requiring substantial new investment, trusting in relative peace within their own market segment. Continued expansion of broadband capacity was on the verge of stalling.
I don't see how that can be viewed as a Prisoner's Dilemma.
Here in Phoenix Google announce Fiber and Cox Communications when crazy rolling out their gigabit offering. Google stopped, Cox pretty much stopped too.
There’s also the theory that Google will eventually provide the consumer service for SpaceX’s internet satellite constellation. Laying cable was way too expensive and won’t be used for most of internet traffic in the future, where low latency is not paramount.
If I have the timing correct, Google led a large financing round of Spacex around the same time as they stopped expanding their Fiber service
In the areas where it was deployed, Google Fiber was wildly successful. And yet, Google threw in the towel. Why?
It _isn’t_ because Google Fiber is was unprofitable. To the contrary, it was profitable _despite_ driving broadband prices way down in its service areas.
It’s because the incumbent ISPs simply have too much power. That power comes in many forms:
etc.If the incumbents are so well protected that one of the most powerful corporations in America found the barriers to entry too burdensome, then how can anyone argue that there is any semblance of a “free market”? What hope is there for startup ISPs?
The incumbents have received hundreds of billions of taxpayer dollars, so they already had a government-assisted head start. And, despite prices being so inflated that they halved overnight wherever Google Fiber rolled out, they are still able to effectively stave off meaningful competition.
Would that be possible in a healthy, competitive marketplace? No. So we need to stop treating it like one, when it clearly isn’t.