FTC and several states sue Frontier Communications regarding internet speeds
reuters.com
reuters.com
This complaint is about Frontier's decrepit DSL-over-copper service, which is the only service they offer to 80% of their customers, and varies widely based on loop distance (to the central office) and line quality. Very few of these customers get more than 20 Mbps. [1]
Hopefully Ars Technica will have a post about this lawsuit that contains more context.
[1]: https://arstechnica.com/information-technology/2020/04/front...
Line attentuation can be a huge problem because a lot of rural homes are far down dirt roads away from infrastructure.
https://www.statista.com/statistics/185532/us-household-dial...
Story time: I worked at a CLEC in the late 90s, and my entry there was in tech support. People would call in and literally scream at me that they couldn't get better that 26.4Kbps from their dialup even though they lived in a brand new subdivision full of expensive houses, so it was obviously due to our incompetence. I'd patiently explain that rather than running new lines to their new lots, SWB had installed a pair gain, and no modem was going to get decent connections through that beast. The sheer fury these customers would direct at SWB after I told them was an awesome and terrifying thing to behold.
My other option is Comcast but for anything over a 20mb upload speed I'd be paying over $240. Even their fiber doesn't get above 20mb up, last I checked, without reaching a pricier package.
I did have Quest DSL in the 90's and loved I could pick my ISP. (Hello Spiretech!)
Worked great, until it didn't and it was always another phone glitching it. Still, for the mid 90's, getting 600kbps in was sweet!
Seems like they were very aggressive about loop distance, not setting realistic expectations.
Now the ACCC (equivalent to the FTC but with teeth) had created this situation by allowing third parties to put their own DSLAMs in the telephone exchanges. In response, that telco (Telstra) let the landline business languish (after all, why invest in a network that competitors have a right to access?) in favour of 3G (and later LTE), which had no such constraints and was much more profitable.
2-10Mbps was good in 2002. Not so much in 2012.
So the NBN's original mission was to replace all this copper with fiber (FTTH) and was going to cost tens of billions. Each household would be guaranteed 100Mbps.
Construction got off to a slow start and after a few years later not much fiber had been laid. The Labor (left-leaning) government lost an election and was replaced by the Coalition (Liberal-National parties; conservative) who had to be different and proposed a mixed-technology model as a cheaper alternative.
Their model meant for most people they were going to get fiber-to-the-node ("FTTN"). This is the last-mile copper using VDSL1/2 (which only works over short distances) and otherwise fiber.
So for $50B+ each household would only be guaranteed 25Mbps. If this saved any money at all I'd be surprised. But that's what Australia has now. And not everyone gets 25Mbps. By network topology some people aren't within 1.2km of the fiber node.
DSL just shouldn't qualify as broadband (of acceptable speed) in 2021.
[1]: https://en.wikipedia.org/wiki/National_Broadband_Network
Here they’ve split the ILEC-equivalent (BT) from the company that is responsible for providing wholesale access to BT and the CLECs and that company owns the physical plant and infrastructure, including last mile (Openreach). BT owns Openreach but it is semi-independent and seems to be heavily regulator-driven to provide as much access to as many people as possible.
Compared to other places I’ve lived, this model seems to be working better than most.
Current streaming codecs are impressive though, no problem streaming 720p YouTube.
Hoping to get StarLink soon, just hope the ping isn't horrendous.
I was very fortunate to find a small local wireless ISP. It’s not fiber to the home, but 10Mbps wireless beats Frontier’s DSL any day.
To top that off, I've heard from my old friends and family, that the lines in some neighborhoods are so overwhelmed by the current load, that they are have begun to turn down new customers because they do not have the infrastructure to support them.
West Virginia and Frontier have a long and troubled history and it seems like it's going to continue for a long time. I can only hope that StarLink is a real and viable solution for people who don't have good internet access.
Some details about WV's past with Frontier https://www.wvpublic.org/government/2020-12-10/frontier-has-...
She got Starlink and now is getting about 80Mbps down and 40Mbps up.
It is a bit pricey at $550 for the unit delivered and $99 a month. But when you don't have any other realistic options...
How do you do this? Starlink is location-locked at the moment. They're awaiting FCC approval for fully-mobile use, and other people have reported that they can't even get service in a fixed location that's more than a few miles from their registered service address.
Doesn't every ISP do this? Isn't that why we have the wording of "up to n Mbps"?
I have had three ISPs in the past 5 years, and only one of them was able to occasionally deliver the speeds that they advertised for the plan during off-peak hours. I even made a twitter account just to let Verizon know, their reps didn't seem to appreciate it: https://twitter.com/VerizonFiosUser
I mean it's great a lawsuit has been created to address this, and I hope they win. But I also am pessimistic that this will lead to meaningful change for the big ISPs. Frontier might be the lowest hanging fruit to go after, but they are all doing this.
On another note, I could see value in a separate certification process that shows the actual average speed delivered for each plan in a given area.
IMO "up to N Mbps" is fair if the ISP routinely delivers, but if they never deliver N Mbps (hey Frontier) then it's just plain fraud.
At best it's misleading. Frontier should adjust the contract and marketing language to reflect its true capabilities. If it can deliver "up to 100 Mbps", but on average/typically can only deliver 50 Mbps, the marketing language and contract should reflect that, e.g., "50 Mbps on average/typically, with bursts up to 100 Mbps possible."
They put a fiber node just out of VDSL range but I have ADSL with two lines at 18 Mbps/s with an application-level load balancer. Each of those lines performs as rated almost all the time.
In the early 0's Frontier struggled with "middle mile" capacity even for customers who only got 1 Mbps DSL. People would have their service turned off with no explanation in my neighborhood, I had gone all the way up the CEO of the company to get help with one of my lines, the truck was followed by some lady in a car who was desperate to get hers fixed. Even when it worked the 1 Mbps DSL like would give more like 30k.
Eventually they fixed it, but around that time they announced they would implement a miniscule 10 GB monthly data cap which they were forced to suspend indefinitely once the public heard about it.
(I understand that upgrading DSL is a challenge, but there is no excuse for not upgrading the middle mile.)
Ok, I worked in telecom and saw the roll out of priority bursting... to me, this makes your data a yes.
Every ISP does in fact over sub.
They put in boosting and over provisioning to stop complaints.
The service would be far more expensive if they had to guarantee sustained speeds, or rather that you had a guaranteed bandwidth potential.
It's not really a factor in this question though. The discussion with Frontier is they sold speeds that were impossible to reach, ever, not that the sold speeds you couldn't reach 100% of the time. Most ISPs do not do this, the vast majority will at least hit advertised speeds regularly even if they aren't guaranteed to at any instant.
And I’ll concede that oversub for DIA is totally different. Surely it’s still possible somewhere but ever less and less likely considering how much larger a 1Gbps fiber pipe is a 56k connection to 15k people at once.
But for actual broadband... I know ISPs over subscribe because I’ve seen it. I’ve been on the calls and in the room when it’s being calculated. Literally everyone does it. Speed boost and overprovisioning are tricks to please customers who are willing to accept “lots of people are online tonight so it’s slow”.
I can't even pull 150mbit at full view of the 5g tower with best connection. That's less than 10% of the 2gbits they sold me and 1/5th of what the support told me I can expect.
Meanwhile my cable provider is providing me with 90% of what they sold me which I find acceptable.
I'm now running 15min automated speedtests and hopefully find someplace I can submit this data eventually.
I hate being lied too. If they told me I would get 150mbit I would have been happy.
No. Most ISPs overprovision. e.g. my Spectrum service is nominally 600Mbit but actually supports 680. This is done because a) users often measure payload throughput vs wire throughput and there's a ~10% packet overhead, and b) to avoid customer complaints at the margin.
I also gained a lot of insight into the tools available to support staff and they varied quite a bit. I even worked on a contract with Frontier for a few months. The service they offered for former Verizon FIOS customers was decent, but DSL customers were treated like crap. Unless a customer went out of their way to make multiple contact attempts and demand to escalate multiple times, they would be largely ignored in any situation that required a truck roll. The company truly did not care and it's obvious nothing has changed (Except maybe for the worse) since I worked that contract.
Of the ISPs I worked with I actually found Centurylink to have the best tools and best service. Personally I've been stuck with comcast for most of the last decade so I've never been the customer but the support staff was generally pretty well trained at the time and they had a lot of information available about the line status, as well as direct interface with the CPE. They were in a transition to off-shore support at the time though, and it was clear that the quality of service provided by the off-shore agents was much lower. They were very clearly trained off a tight script and were woefully incapable of handling situations that didn't fit the script.
Every ISP oversubscribes. OP is correct.
Overprovision is a different thing. You are talking about getting 680 so that when you speedtest, you have NOTHING to complain about to them. This is basically part of their boost and oversub plan, where it isn’t hurting them so they toss you a bone to preempt speed complaints.
What you aren’t told is that Spectrum also does short term speed boosting. For 30S-1min you’ll get 680, but it will fall in PRIORITY after that. This also conveniently helps speed tests. At 3am, you’ll get a solid 680 real world. At 6pm, you’re more likely to get a sustained 680 for a bit, then fall to 400 or so. You need a transfer at max speed to show this. Newsgroups are good for maxing a connection and showing the fall-off decline of a boosted connection. Spectrum only started this in my area a year ago or so. After an initially great move to DOCSIS 3.1.
Oversubscribing means that if Spectrum has 2TB of bandwidth, they’ll sell 3TB knowing that all their customers won’t concurrently use it. It would be a statistically unlikelihood. TRUST ME, they do the math on this.
If you add all the 600Mbps users up, you’ll get to a number they can not possibly support. This is OK.
What the issue that Frontier seems to be in trouble with is they went way too far, and couldn’t hit the base numbers they were offering.
But YES, every ISP out there I ever felt with while in telecom does in fact oversubscribe.
Is that really how it works?
I was working for a small-town ISP and our approach was: Boss link A is at 50% and will fill up within a month. Shall upgrade the converters?
I mean, you could argue that we did math, i.e., linear prediction on statistical aggregate values, but this was not synced with sales or marketing, rather done out-of-band.
I'm talking here about access network and core network, not last mile.
It sounds from your other comment it was a smaller ISP you were at. We definitely had the data from NOC and Sales and watched them figure out what hardware they would need for the next few quarters based on their oversubscribe estimations.
Edit: lol; I worked in telecom. I’m sorry first hand experiences make some of you so mad.
With Frontier, you're not talking "I pay for 100Mbps and sometimes I only get 80Mbps.".
You're talking "I pay for 5Mbps and I've never seen 5Mbps. The best speed I've ever recorded was 1Mbps, and my connection regularly fluctuates between 0.1 and 0.3 Mbps even during off-peak."
This is a cop accepting that people will do 70mph on a 60mph road, but not 120mph.
What comparison does this possibly have to law enforcement.
You cannot advertise 12, then sign a contract that says 12 on the front, but has “go fuck yourself, you won’t ever see 12 in your life on this plan”. That’s false advertising.
If you do that, you run a real risk of finding yourself the subject of a lawsuit filed by the United States Federal Government, and potentially state Attorneys General
I once spoke to a customer service support individual who said I was "lucky" to be receiving 60 mbps of download speed.
You can put whatever you want in a contract, but the law isn't a dumb computer mechanically executing instructions. It's pretty obvious when you are willing to charge people twice as much for "up to" twice the speed while providing an identical service that you are not presenting your products accurately.
> What comparison does this possibly have to law enforcement.
The comment I replied to literally said "Sounds like the ISP equivalent of 'when everyone is breaking the speed limit all the time police have justification to pull over any individual car they want'". It was the first line. If a comment seems confusing in context, it might help to _read_ the context.
Oh no, it isn't confusing at all. You encouraged a spurious comparison to a baseless allegation. There is no parallel to law enforcement, whatsoever.
No. At least in NYC, every provider I've ever had (several) normally achieved the advertised speeds during the daytime mostly, and overnight (e.g. 3 am).
But it might go down 25% or even 50% in the evenings once everyone gets home and starts watching Netflix, at certain times on weekends, etc.
(Obviously COVID has changed these patterns a bit.)
That's always felt generally reasonable to me for residential contracts -- that the capacity is there except during peak residential usage or temporary maintenance.
"Up to" with regard to internet speeds generally means you will reach that capacity at non-peak times.
If you can't regularly achieve max capacity at 3 am, then the advertised speeds are fraudulent. After all, bringing this to its logical conclusion, companies can't advertise "up to infinite speed" even though that's "mathematically true".
One friend that was on Frontier's DSL out in the boonies had a 5Mbps plan. At peak times, he'd be lucky to see 0.2Mbps. (And he'd be lucky to get it up to 1Mbps the rest of the time...)
It's really hard to justify providing 5% of the "up to" speed at peak times as reasonable in any circumstance.
In his case, Frontier's answer was to tell him to upgrade to their ~10Mbps plan if his speed wasn't enough. Never mind that they already weren't delivering what he paid for. After several complaints to the AG's office and other regulators, he eventually was in touch with an engineer that just told him flat out that the node he was on was already oversold by many hundreds of percent and there was just no way he'd ever see any promised speed.
So it wasn't even a case of Frontier thinking they had a realistic chance of ever delivering the speed they had sold and dealing with fluctuations in demand. They very clearly sold a product that they never expected to deliver.
The basic history of Frontier in my area is that they bought up all of Verizon's old copper lines and have done exactly nothing to improve the capability. They don't even know what the capabilities ARE. Areas are either oversubscribed or so far from nodes that the signal is worthless. They're trying to squeeze every last cent out of rural houses with no other options and letting the underlying infrastructure slowly fail.
I ended up using a pretty performant (~50Mbit) LTE connection, hacked AT&T hotspot and yagi antenna for 2 years. Finally, as part of a rural broadband grant program New York State got Altice to come in-- probably in exchange for a monopoly in a more attractive part of the state. Now I have 300Mbit cable for $100/mo. Altice (aka Optimum) is also a terrible company, but Frontier still sends me postcards claiming to be "the best internet in <my town>".
These guys are selling a product that doesn't work. They deserve to be sued and I'm disappointed that New York State isn't involved.
Mobile internet can be a bit variable depending on time of day and the provider, but still way better than DSL for me, I get anywhere from 15 to 100Mbit throughout the day, although this level of stability was only achievable using a Cat20 LTE router (Netgear Nighthawk M2 with a couple small external antennas). They probably seem a bit expensive but it's worth the investment, not only for throughput stability but reducing packet loss.
Just don't ever buy mobile contracts for this stuff, at least not before trying out a PAYG sim, you need empirical testing to find out which provider is viable. I attempted to use open cell maps to figure out which provider would give best carrier aggregation in theory, but you don't always connect to the masts you expect, and even then it's reliant on back-haul availability and subscriber contention ratio anyway. In the UK a couple of the main mobile networks have unlimited 1month rolling contracts under different brand names now... so real competition is now possible!
Luckily, in my region Frontier was sold to Ziply last year, who immediately started investing in ditching the old copper wires and now they offer symmetrical gigabit fiber here, so I switched back and have never been happier.
Based on what I've read over on the Ziply subreddit, Frontier's network was in poor shape when Ziply took over. For example: https://www.reddit.com/r/ZiplyFiber/comments/kw56u4/does_zip...
Mini Rant:
They went through bankruptcy, struggling to make ends meet, no bonuses, so all good employees left within first few years, no investors wanting to risk with installing or maintaining physical cable, and people, especially in this forum with hatred of ISP's, have no idea how expensive it is to maintain cable lines to every house in US, yet they expect GbPS for 39.99. 39.99 doesn't even cover truck costs for 5 years when somebody drives over or cuts fiber line on the street or digs to plant a tree, forget about any profit or return for the company.
I don't care if they had good customer service, or were working on a shoestring budget. If they sold something they couldn't do, then that's a crime.
Infrastructure is an investment. You deploy infrastructure and then have customers pay for it over the next decade. Unfortunately, Frontier has the highest customer churn in the industry so the cash flow is really not great. Their challenge is to figure out how to execute that 2nd investment even though the first one never paid off. Over the past few months they completed their bankruptcy and shed a ton of debt, so they may actually be able to pull it off.
Simultaneously, there are new challengers in the field like Google Fiber, ImOn, MetroNet, and municipal initiatives who are all making their first investment with fiber... and they all offer pretty good pricing.
Well, to be fair, the reason they chose DSL is so they wouldn't have to run infra to every house - POTS lines were already there.
In the recent past, when Frontier had extra cash, they used it to offer extra dividends to stock holders rather than building infrastructure.
So please stop with the excuses.
edit: Asking because it would be cool if you could expand a little bit on the troubles faced by Frontier or similar ISPs, if you have any info there.
>Please don't post insinuations about astroturfing, shilling, brigading, foreign agents and the like. It degrades discussion and is usually mistaken.
What does your comment add?
Side note - it's funny, in a wry way, that my question can be simultaneously asked in (what felt to me like) such an innocent way and interpreted in such a malevolent way when you view it through the lens of "standard HN snark".
I'm not sure I understand what you are saying. Surely you understand that it's 39.99 a /month/ and a cut cable normally affects more than 1 household (so it's 39.99 times households affected). Also, it seems like you are conveniently leaving out the billions that the US government have given to ISPs to improve/build their networks. Bad/failing ISPs have no one to blame but themselves. I don't hate all ISPs but I do hate the ones that are scummy (which the big players all are). I gladly pay ~$100/mo to MetroNet for gigabit, they are about as no-nonsense as you can get in my book.
This hand waving and calling companies "bad" because they failed financially while at the same time complaining about monthly fee or speed or support is oxymoronic.
Try this as a experiment, call a local insured commercial digging company and get quote to lay cable from alley to garage under the driveway/pavement WITHOUT damaging it or any other utility lines and restore all plants/grass/sprinklers after they are done. Ask them quote for the neighborhood, oh if there is any HOA get some lawyers to play HOA politics and write up documents to convince HOA to allow digging/construction.
Now try to make that money back at average rate of $40/month while maintaining those lines, paying electric company for renting/sharing the posts/properties.
>I gladly pay ~$100/mo to MetroNet for gigabit, they are about as no-nonsense as you can get in my book.
You is NOT everybody on the street, not even remotely are people willing to pay $60/month for gigabit, even if you lay gigabit line, people in general are only wanting to pay $39.99 or $49.99 and get 50 or 100Mbps.
May 27, 2020: Court signs off on $38M in executive bonuses at Frontier. (Source: https://www.channelfutures.com/telephony-uc-collaboration/fr...)
SEVEN WEEKS LATER, July 16, 2020: Frontier seeks, and is granted an ADDITIONAL $25M in executive bonuses. (Source: https://www.law360.com/articles/1292694/frontier-wins-ok-for...)
Have you ever paid for internet in the US? GbPS is more like $80+ first of all (especially with fees), second both federal and state governments have given Frontier alone nearly $2 BILLION dollars to get their network up to snuff.
https://ilsr.org/fact-sheet-frontier-communications-has-fail...
I am not surprised to learn they have had financial problems.
Is there anything technical preventing companies from offering dedicated speeds and lines at a premium at residential neighborhoods other than pure economics?
If so, I think phone users will have an interest in justice.
[1]https://arstechnica.com/tech-policy/2019/05/ajit-pai-refuses...
Frontier was so much better than Spectrum — pings to 8.8.8.8 are consistently 6-8ms, compared to Spectrum at 10-30ms. And Frontier download speeds are 500Mbps-800Mbps, significantly higher than Spectrum.
The installation was also more competent and organized.
I'd guess there's lots of regional variation though.
I plug my Google WiFi (yes I work at Google) directly into the modem and ignore the router they gave me entirely. It's real nice not having to deal with that huge box.
One thing that was unexpectedly bad was their routing. It was just atrocious. Playing any online game would see ping constantly jump between 50 to 180 ms as their network did... whatever the fuck it wanted to do. I had to use a VPN while gaming just to get a stable ping. Getting the packets out of frontier's network ASAP was the most important thing to have a usable experience. I've never seen anything like it before or since.
I tried to trace where the issue was. My best guess was the issue had to do with their uplink to the rest of the Internet. At times in the day it would get saturated because they didn't provide enough bandwidth at that point. There was a bottleneck.
At the peak of Zoom school, I had 3 kids on zoom calls, my wife and I both on zoom calls, and the 2 year old watching Netflix all at the same time without even a blip.
I'm 90% certain the gigabit download isn't that important, but synchronous upload and download speeds that make all the difference.
ISPs that have many locations around the US are routed differently to the broader Internet and can have different footprints in different locations.
I’m not affiliated with the company, but I am a customer. We switched from Frontier to T-mobile and went from 5mbps to 150mbps for about the same monthly rate, $50/mo.
I distinctly recall that at the time I was looking to rent somewhere less pricy and literally crossing outside of the city lines relegated someone to living in 'the internet ghetto' of ADSL at 7mbit down and less than one up if you were lucky, or cable with outrageous rates and gigabit deploying years after it became available in the city of Seattle. There were a couple expensive apartment complexes served by Wave broadband, which had acquired an earlier company, 'condo internet'.
I live about 50 miles from Seattle. 120 from Portland. My TV is Seattle-based.
Basketball games for the Portland Trail Blazers are unwatchable, even with League Pass, because I cheerfully get told the game is blacked out and to "watch on my local Portland channel".
Prior to that, I was on Frontier who took over Verizon fiber, which was setup at my house, but unused due to their insistence I pay a very unreasonable setup fee. Was $600, and no way. It was up, ready, nothing needed. Pretty sure they did a free move deal for the people who lived there prior thinking they would get money from me too. Nope.
I used a 3G dongle for about a year standoff with Verizon. Also had a Moto droid with Verizon unlimited running the house for a few months too. Was a work phone with that great plan baked in.
Frontier waived all that and I bought TV plus Internet. (Fam wanted TV)
Frankly, I loved Frontier!
Tried to setup auto pay and it would break. This happened a few times, so I left it broken figuring someone would call and we could sort it out then.
8 months go by, and service stops.
I call, expecting a big bill and terms.
The service woman and I had a chat about the back end mess and it was truly spectacular. I asked her why she worked there and she said the crew was the best, systems not so much. Fair enough.
It was broken, but they laughed a lot. Love that!
Anyway, she said it would be easier if I were a new customer. No big bill, no terms, just start over. And new customers get a free year of Amazon Prime.
I asked her is she serious? Answered yes and this spoke to why they laughed a lot. The system was broken, but they had wide latitude to take care of people because it was broken enough the higher ups rarely could sort it out otherwise!
She was a hoot! Great chat while she got me all setup on the new system, autopay working just fine.
TL;DR: Payment system failed and I was shut off for non payment and got free Prime and better service for less money despite being 8 months over due!
Other than that, I love an ISP I never hear from, unless I really need to. Never heard from Frontier. And it worked just great. Seriously fast, basically no downtime.
The only negative is the old Verizon fiber interface and its pain in the ass battery. I finally just wired a power supply to the battery so it would shut up and I no longer change batteries.
I do not hear from Ziply today either, and it works great, and cost makes great sense for everyone.
Never do this in the US. You need to proactively call them to resolve the situation. The standard thing for nonpayment -- which maybe Frontier didn't have working, but you're taking a real risk -- is for these to go to a collections agency, and it will go on your credit record, which affects your future ability to get credit for large loans of various kinds.
Having credit destroyed and basically trading my home for my wife due to both her getting colon problems, and an employer who thought a trip to Spain mattered more than insuring people all left me in a zero fucks to give position at that time, so I let it play out. Unwise? Inadvisable? Yup. :D
(We really do need significant health care reform here, just saying. Mid 6 figure events are no joke)
There were some prior calls too. All initiated by me. Do this, do that, oh you want paperless billing? Sure! And on it goes.
I'm not sure if merging is the answer for them to succeed, but the talk in this space has been very much in that direction. What remains to be seen is how much of an impact Starlink will have on their customers. In a way in might be helpful, allowing them to rid themselves of their most expensive to service customers. But if Starlink becomes popular in small towns rather than just rural areas, that could eat into the easier to service and profit from customer base of these companies.
https://www.treasurydirect.gov/govt/reports/pd/gift/gift.htm