As two of Canada's biggest Internet service providers, Bell Canada and Rogers Communications are fierce rivals that frequently battle for the same customers. That marketplace fight rarely spills into the courtroom, yet my weekly technology law column (Toronto Star version, Ottawa Citizen version, homepage version) notes that last month a Rogers advertising campaign prompted Bell to file a $50 million dollar lawsuit. The result was an end to the campaign and evidence both companies over-promise the speed of their Internet services.
The case began when Rogers launched a direct mail and Internet ad campaign called "Check Your Speed." The campaign warned users the Internet services "you are paying for may not be what you're getting" and encouraged them to test their connection with an independent third party. The campaign unsurprisingly offered Rogers services as an alternative, promising a "reliable speed every time you connect." Just days after the launch, Bell filed suit, arguing in court documents violations of the Trade-Mark Act, Competition Act, along with various torts. The company sought $50 million in general damages, $1 million in punitive damages, and an injunction blocking Rogers from continuing with its campaign.
Two days later, Rogers dropped the third party testing feature. Rather than using a fully independent third party service, Rogers had used a server located in Seattle, Washington to run its tests. The court found that the distance between users in Ontario and the speed test server in Washington might help account for slower speeds. Even more telling was the evidence that placed the spotlight on a Canadian industry practice of advertising the maximum or "up to" speeds for customers, rather than minimum or actual speeds that customers typically obtain. The Rogers campaign was effectively premised on this discrepancy since it encouraged users to check their speeds where they would undoubtedly learn their typical speeds were lower than those promised by their ISP.
The same holds true for Rogers, however. Under cross-examination, a Rogers witness was asked about the Rogers service: "if he [the customer] runs the test, and it tells him that his Internet connection speed is less than 10 megabits per second, is he still getting the Internet speed that he's paying for? "
The response? "The Internet speed he's paying for is "up to," as is industry practice. So I believe the answer is yes. We sell an "up to" service, as is industry practice."
It is this "industry practice" that deserves far closer scrutiny from Canada's telecommunications and competition regulators. While the Canadian Radio-television and Telecommunications Commission has done little to address the issue, its counterparts in the United Kingdom and Australia have taken action.
Late last month, Ofcom, the British telecommunications regulator, released the results of a study that found more than 50 per cent of broadband users in that country are receiving less than half the speed promised by their providers. The Australian Competition and Consumer Commission addressed this issue several years ago, with guidelines for ISPs designed to counter misleading "up to" speed claims.
The Canadian practices are particularly deceiving since ISPs have also been slow to disclose their traffic management practices, which may often result in deliberate slowdowns for certain applications. Rogers recently admitted it charges tiered pricing for faster upload speeds but that all tiers are throttled to the same speed when using peer-to-peer applications.
As Canada's ISPs battle in court over which company is more deceptive, it is time for Canadian regulators to step in by conducting their own tests on promised speeds and establish clear requirements that bring truth back to ISP advertising.
“Up To” is Relative
I feel that Rogers is in the Right here. Although both parties sell an “up to” service, with Rogers users usually get at least 90% of the maximum while Bell users often get less than 50%.
As someone who provisions and supports hundreds of small business pipes, I see this discrepancy every day.
A.
“Up to”
I can’t speak to the service provided by either, as neither offers high speed in my area (I use dial-up). Thus while I have a service that offers “up to” 56 kbps, the reality of the matter is that, due to line length and noise, I have only once ever seen a speed faster than 26.4 kbps. While neither are exactly lying, both are fudging the truth some… of course, you’ll never see them advertise real numbers. Part of this is that the real numbers you’d see are a function of what you do. Email and standard web browsing bandwidth is burst traffic. One the other hand P2P and gaming is more sustained bandwidth and is therefore more liable to be throttled.
Bell and Rogers Square Off Over Internet Speed Claims
The speed thing and the up to “industry standard” is part of the smoke and mirrors game both of them participate in.
They are truly just fighting over lipstick colours on the pigs that they both sell.
If one takes the maximum download usage amount per month and divides it by the download speed that is a good indication of the amount of time you can fully utilize the connection. That is the real dirty secret of both of them, that tells the real bandwidth story.
When you can only use 60 gallons of water a month how much does the tap size really matter?
Kitchen tap or Fire Hydrant? Perhaps the reality of the situation is more suited to an ornamental table fountain.
Editor, DSL Prime
Michael
These are real issues I’ve been tracking around the world. Let me know if you need datapoints, especially about DOCSIS 3.0.
Turns out most networks, including Virgin cable in the U.K., my own Verizon and TIme Warner, and most of the U.S. cablecos are doing very well on performance, delivering 80-90% of the “up to speed” – except for DSL on long loops.
This makes sense if you think of the debates around “congestion,” which is rare on Bell Canada, Rogers, and most other carriers. IF the network is not congested, there is no reason it can’t deliver the promised speed (if the other side can serve it that fast.)
I’ve been inside several major networks, including one in Canada, looking at congestion that might affect throughput. It is amazingly rare given the rhetoric.
I learn from you and always happy to share files and information. db
Proper bandwidth testing
The speed debate is one that has existed since we first deployed DSL back in 1999 while I worked at a Canadian CLEC. The speeds of the connection are between the DSL modem and the DSLAM, also known as “sync rate”. If you were close to the central office and had a clean phone line, your sync rate would be 3Mbit x 800Kbit. If you were farther away, you might see less than that. This is where the “up to” came from.
It would be easy for the carriers to tell DSL customers what their sync rate is and many DSL modems (for example, Netopias) will tell you the sync rate even.
I think what people (consumers) fail to realize is that sync rate isn’t a measure of true “bandwidth” capabilities. Once you get past the DSLAM, there’s a whole lotta network a consumer’s packets must pass through before hitting Facebook, Google, Flickr, etc.
Even if (and this is a big ‘if’) The carriers gave a consumer full bandwidth all the way to Facebook/Google/Flickr/etc service, a number of other factors kick in:
– Does Facebook, with its millions of users have enough bandwidth to serve up the information at full speed?
– Do Flickr’s servers have the ability to respond quickly enough to reply using the full bandwidth?
– Carriers often do not have direct connections to companies like Google, Facebook, Flickr, etc and so packets traverse a network that isn’t owned by the carriers. This is why peering is important (Rogers peers at TorIX for example, while Bell does not).
The peering thing is a big deal for the same reason that Rogers’ independent testing wasn’t valid because they were testing from a server in Seattle. If Rogers had better peering connections, the testing results could be in their favour.
Here are some valid tests:
1. Consumers could be given access to a ‘speed test server’ that’s located close to the carrier’s edge network (this is the point before the Internet traffic leaves the carrier-owned network). This test would theoretically show the consumer how fast their connection is ON THE CARRIER’S NETWORK. Not that useful, but still a good test.
2. This test likely won’t happen, but consumers could test against speed test servers on a service’s network. This means that Google/Flickr/Facebook/etc could let a user test their speed all the way to the service they’re actually interested in using.
What I guess I’m trying to say is that speed testing is a tricky process and I don’t think either carrier wants to go through the independent testing required. Even if they did, getting test points setup on networks outside of their own would prove difficult.
In the end, as a telecom professional in the industry for more than 15 years, I’m confident that Rogers would would win the bandwidth speed test debates against Bell Canada, especially due to their efforts of setting up peering arrangements at places such as TorIX.
Perhaps future online services (especially media heavy ones) will implement speed testing servers (like I mentioned in #2) and consumers can make a decision then as to which carrier they want based on those results.
Editor, DSL Prime
Michael
These are real issues I’ve been tracking around the world. Let me know if you need datapoints, especially about DOCSIS 3.0.
Turns out most networks, including Virgin cable in the U.K., my own Verizon and TIme Warner, and most of the U.S. cablecos are doing very well on performance, delivering 80-90% of the “up to speed” – except for DSL on long loops.
This makes sense if you think of the debates around “congestion,” which is rare on Bell Canada, Rogers, and most other carriers. IF the network is not congested, there is no reason it can’t deliver the promised speed (if the other side can serve it that fast.)
I’ve been inside several major networks, including one in Canada, looking at congestion that might affect throughput. It is amazingly rare given the rhetoric.
I learn from you and always happy to share files and information. db
Fought with Bell for years
I’ve had a problem with Bell for four years where I was paying for the 5 Mbit service but never getting better than 1 Mbit speeds, ever. I had numerous phone conversations with Bell tech’s trying to resolve the issue to no avail. Finally a tech was sent to the house where he measured the signal strength only to find my house was too far from the central office. I feel that Bell was charging me for a service that they were incapable of providing and I wanted to get a credit from them on future bills. Unfortunately I had no luck getting anything from them because of the “up to” clause.
Instead I canceled all of my services from Bell and switched to Rogers. While Rogers is certainly not a perfect company I have found their internet services at least delivers what is advertised.
Time for Regulation
I think it goes to show that Government has to step in here with laws. I am not a fan of regulation, but no other industry can sell a service that they ‘know’ you will not get. Up to is the biggest lie. Even the current system could be changed from Up to 10 Mbps to 15 Mbps with no changes, since you will not get it anyhow, but they only have to show that it is ‘possible’.
I think it is time that the marketing was forces to give a lowest speed value, and then the providers must honour that. If I am only ever going to get 6 Mbps on my 10 Mbps service, then they should only be able to sell it as 6 Mbps or lower, and then ‘over-deliver’ and I should not have bandwidth under that mark.
Up to Speeds?
Imagine if all cars sold were tagged with “UP TO 180KPH”. My Volvo can do UP TO 240KPH. Not that you’d ever drive that fast. It’s ludricist for ISP’s to be selling ‘UP TO’ speeds. They should be billing on what they can deliver on average speed rates. No different that cars are sold on what the Average Fuel consumption rates are. Not maximum speeds. I don’t buy a car because it can do 240KPH. I buy it because I know it can get 7.7L/100km (30MPG) on the highway at 90kph.
Shouldn’t ISP’s treat this the same way?
I’ve been using Telus ADSL for almost 3 years. I’ve not had complaints until recently when certain downloads crawl or die. These are generally podcast downloads from BBC or American Public Radio.
Is it possible that Telus suddenly started throttling that sort of traffic? Is there someway I can determine the true situation.
Latency is just as big a problem
I’m a gamer and need low latency or ping values. The ISP’s are playing some interesting tricks here as well. For example, my connection is ALWAYS routed through Toronto, regardless of the destination. So if I’m connecting from Sarnia, ON to Chicago for example, my connection goes several servers out of its way to get there. This is basically doubling my latency.
Presumably they want to route me through some sort of filtering or who knows what at their main server in Toronto before sending me on my way.
once again, CRTC is to blame for doin’ Nuthin !!!
Bell and Rogers both really are terrible performance wise, at ALL levels compared to what they used to be like.
Anyone remember the “Granfatered DSL…” ? 🙂 In its day the best in the world.
But Now …. ummm actually, when I think of Bell and Rogers today, they are more like that SNL show “The Ambiguously ___ DUO” -that was funny though…
Except for the fact that Bell and Rogers seem to air out their little “marital” spats publicly -which is not so funny.
And the CRTC act like embarrASSed parents ?
“…. What is everybody looking at…” ?
-“NUTHING” !!!
😉
cheers.
Re: Latency is just as big a problem
Hi Chris:
I’m with Acanac (who basically gets their DSL lines from Bell) I’d like to add that there is definitely something weird, “illegal” filtering going on as well.
“Voip”ing is horrible now. -unless of course you’re paying Bell/Rogers dearly for it.
I remember when “skype” for example you could talk for hours and it would rarely drop.
Now though, its like every 10 – 20 minutes skype/voip calls drop all the time.
This severly hampers online gaming as well since many groups like to voip/audio chat while playing…
Its not my freinds online in europe or U.S. thats dropping a lot -its my (Canadian)ISP up-the-line that’s causing it.
ya I realize Skype itself ain’t exactly the best VOIP app out there anymore, but unfortunately a lot of people use it.
-I tried Ekiga, mumble/murmur, … same thing.
Maybe chock up the “weird” filtering to
1./ its “poor” management/networking technology on the (Canadian) corporate ISP’s part.
2./ “They” obviously don’t want Canadians “skype”ing, or whatever -‘casue they’re losing money obviously.
nothing
well try teksavvy
It is an Internet Service Provider, whose focus is to make technology work for us.
And so far they have delivered on what they promised
Gaming unveils speeds
Not until I started online gaming did I realize I was being ripped off by our ISP…it’s a real shame…
teksavvy
Anyone Saying teksavvy is better then bell should know that they use bell’s infastructure. same lines same packetfiltering bs.
mind you thier teck support is 1000x better then bell.