The bigger swindle is that with most ISPs 10 minutes of downloading will see your bandwidth chopped to a fraction of whatever it was advertised at.
It will come as no surprise to regular readers that 90 per cent of UK consumers are confused by broadband advertising - we'd have to assume the missing ten per cent are marketing bods for ISPs. Virgin Media is to publish "typical" speeds for customers on its different packages, having found that 90 per cent of those it surveyed …
The bigger swindle is that with most ISPs 10 minutes of downloading will see your bandwidth chopped to a fraction of whatever it was advertised at.
"but the likes of BT will say it is unfair to compare a fibre-optic network with its ageing copper network..."
If they did say this, then well, isn't there an obvious answer? :|
Don't know about anyone else, but I'd consider that absolutely fair.
If you can't compete with Virgins fibre network, then stop claiming on all of your adverts that your 'aging' network provides 'upto 20Mb/s'.
Neither VM or BT run a truly fibre network. Both of them terminate using metallic cabling.
*All of which will have been carefully selected to not be in oversubscribed areas.
Last time I ended up in an oversubscribed area with VM, my speed was 2kb/s on the 10Mb connection.
Whilst Virgin are to be applauded for adopting the typical speed approach, I have a few reservations about companies doing their own monitoring - their own blurb states that it's the typical speed attained by 66% of their customers - thus eliminating the 33% that could drastically lower the "typical speed" - e.g. me. I rarely reach anything above 5Mbps on my 10Mbps connection, but Iknow that I'm in a minority, my friends who are on Virgin Broadband get extremely good connections, I'm just in an area that was wired very early for cable TV - it's a conservation area of Edinburgh, dishes were not allowed, so the cable company (I can't remember which one, it's been through so many variants, eventually being Telewest..) saw an opportunity, and a captive market, and the wiring is a tad antiquated. I will say though that it is very, very reliable, I've only had one outage of my broadband, lasting about 2 hours, in the last year, so I won't mention the traffic shaping...oops..
Instead of allowing a self-selected basket, Ofcom should be more pro-active in insisting that the typical speeds must come from an overall group of customers, otherwise it's not that much more accurate than the "up to" figures normally quoted.
I'm not sure what you're after here, you've already said you think you are in the minority, so you're not a "typical" customer.
If (say) in a sample nine people are getting 10Mbps and one is getting 5Mbps, that still makes the average 9.5Mbps. Even if you switch those figures to 7@10 and 3@5, you still only drop the average to 8.5 with most of them getting more than that and you still don't have a figure that really represents what the lower speeds are.
So, if you were looking for a BB connection would you prefer the advertising to say "10Mpbs", "Up to 10Mbps, typically 8.5Mbps", "10Mbps, two thirds of customers get at least 8.5Mps" or "10Mbps, a third of customers will get half that".
I think I can spot which one won't sell them very many connections.
What might be of more use would be somewhere you could find out what the typical (or average) speeds in your area are before you sign up, then you would have known that although most people get >8.5Mpbs, you would be more likely to get <5 because of your location.
Half of our customers will get below average speeds?
>What might be of more use would be somewhere you could find out what the typical (or average) speeds in your area are before you sign up
That's what speed checkers are for. Not terribly accurate (and in some cases way off the mark) but most ISPs at least encourage you to use them before signing up.
I think you'll find that it should be : "Half our customers will get below *median* speeds"
Well median is an average measure as well as mean and mode.
And why median? Isn't broadband speed a continuous variable?
Not to mention Arithmetic, Geometric and Harmonic means.
I wonder which mean they mean when they talk about 'typical' speeds?
"Virgin gets some credit for this increased transparency, but the likes of BT will say it is unfair to compare a fibre-optic network with its ageing copper network and wider, rural, spread"
Why? It's their fault they didn't upgrade their crappy network sooner! Virgin are still an option for the consumer, so surely they should be compared against their rivals.
... by broadband (or indeed any other) product advertising.
Advertising is merely a form of dissembling to try and get you to part with your wedge in a particular way and usually the more advertised a product is, the more likely it is that there's something else better on the market.
Mine's the one with the invoice from a small, friendly, well-performing unadvertised ISP in the pocket.
If the ISPs made cars I would fully expect them to build Trabants, but with speedometers that have labels up to 300 MPH. if they then followed their internet marketing strategy they would sell those cars as "UP to 300MPH *" and "SEATS FOR 22 PASSENGERS **
* Indicated maximum speeds should not be taken as a promise that you can actually attain this speed.
** Subject to our fair usage policy which limits the number of people in a vehicle to 4
Anyone that buys products solely on what an advert told them deserves what they get.
Every sign-up page I've seen has made it clear that xDSL is line length dependant and most invite you to enter your phone number to get a speed estimate. FFS - it's not rocket science. I have never seen anything seriously wrong with 'up to' in general. How else do you describe xDSL?
The only aspect of speed based advertising that is dodgy is the use of connection speed rather than throughput. Unfortunately if people are too stupid to understand the line length issue then they don't stand a cat in hells chance of understanding about TCP/IP overheads or profiling.
That's pretty clear. Can we go back to arguing over the definition of 'unlimited'?
It's all very well saying that ISPs are the work of Beelzebub for using the "up to" description of their speeds but I have not heard a better one.
When you sign up you get an opportunity to view what speed you are likely to get in the real world. Don't see the problem.
"Unlimited" and "fair use policies" are different stories altogether.
If my experience of using a Virgin 'fibre optic' connection at about 5.30 pm a couple of years back is going to be used then 4mb should be advertised as 56k ! At one point I actually dialled up on my spare laptop and it was faster than the cable connection....
Surely the typical speed would be the modal average, the speed achieved by the most users.
Problem with Virgin is that if you're local node is running within its limits then your speed will be very close to the advertised rate. If you're unlucky and your local node is over subscribed then you'll get nowhere near.
maybe the median should be used - the speed where half the punters do better and half do worse. That approach works fine for Virgin's *fixed speed* products where you are effectively measuring their congestion levels / performance. The modal speed could be all over the shop from one day to the next.
How to handle rate adaptive products like BT's xDSL ? Any sort of average is fairly useless without reference to the line capability - if the average 'up to 8M' user is getting 4M of throughput that doesn't mean user A on a long line will get more than 0.5 or user B on a short line will get less than 6M. It doesn't tell you anything about an ISPs variability from peak to off-peak.
So perhaps the right approach for xDSL is to report the throughput as a % of the sync speed, to make it comparable with VM fixed speed products. 6M sync, 4M actual, result = 66%. The samknows boxes attempt this by measuring peak speed achieved at off-peak hours.
"Petter said .... BT was busy building its 21CN network."
I thought that BT had pretty much ground to a halt on any new 21CN Switch 'upgrades' since last year?
Instead of selling packages based on notional speeds with obscure fair-use clauses and traffic shaping, ISPs should be made to sell based on actual usage. If your antiquated copper wire only gives you 2Mb/s, why should you pay for a 10Mb/s service you will never receive? If you need low latency for gaming or want to download huge chunks of the internet at peak time, you should be able to have it, for a price that's clear and proportionate.
Internet access supposedly regarded as being a basic need, just like electricity. Perhaps it's time it was billed for in the same way?
Saying all that, I'm generally satisfied with my VM cable broadband.
>If your antiquated copper wire only gives you 2Mb/s, why should you pay for a 10Mb/s service you will never receive?
Because unfortunately the cost to the ISP is not related to your connection speed. It costs the same to provide a 20Mb/s connection as it does 2Mb/s. The cost comes from how and when you use it.
Someone with a 2Mb/s connection who runs it flat out in the evening costs more to the ISP than someone with a 20Mb/s connection who runs it flat out in the early hours of the morning.
A higher speed means it's easier to be a PITA but it's not implicit. All the ISP's costs are in the exchange backhaul, core network and transits. For most users those work out pretty much the same regardless of connection speed.
"John Petter, managing director of BT’s Consumer Division, said: “Averages are no more relevant to customers than maximum speed...The solution to consumer confusion is explaining clearly what speeds customers can expect on their individual lines when they sign-up."
When I set up O2 broadband in Leeds last year, that's exactly what they did. They noticed I was a long way from the exchange and called me to say my DSL connection would be rubbish - and asked if I still wanted to go ahead with the setup. IMO, all ISPs need to do this.
I just switched to them and whentalking to the rep he said "adsl2 is capabail of delivering up to 20mb but fi we put in your phone number we will see what you are actuley likely to get" witch turened out to be 13mb I will see what I actuley get when it gets set up and there was no limites of mention of trafic shaping in there fair use policey
That statement about average speed being no better than max speed in adverts is complete bollocks. It still not a very good way of measuring it, but is a damn sight better than a theoretical max speed. Best would be to quote both "Upto Xmbs but typically Ymbs" etc.
Virgin are on the right lines, but this should be policed by Ofcom not the ISP (but we know that won't happen)
'Typical' is no use. What is a 'typical' telephone line? Do you have one? I've no idea if I have. You can't even say that a typical telephone line is made of copper since some are aluminium. Others have a different gauge.
That would be like quoting the cost of a 'typical' bag of groceries. Almost impossible to calculate and utterly pointless information when you have a value.
No. 'Up to' is perfectly adequate as long as people realise that it's based on their line and won't normally change.
I mean they have taken fairly well-known words like: "unlimited" for example, and completely changed its meaning to be "extremely limited". And all they have to do to make that legal is to add a little asterisk after it.
The mystery is how they have been allowed to get away with such blatant mis-selling and false advertising for so long, when so many people have told the regulators in no uncertain terms that this has been and still is, happening.
Already had one of these flyers through the post from VM, comparing "average" speeds across several ISPs.
To be honest, I'm surprised my area isn't over-subscribed - I live in a flat, the local council (still) owns the building and the tenants hold leases to the flats. The council discourages satellite dishes so I'd guess there are a _lot_ of cable users - which means VM... and yet I still get average (or better) speeds on my 20meg line (normally 18+ Mb/s).
There are problems with VM of course - but the speed of the broadband has never been one of them for me.
I don't quite get the argument that "it's not fair to compare fibre-optic to copper" - from the point of view of the end-user (in a cable area), what difference does it make? Is VM consistently faster than its main competitors? Yes. End of story... for 99% of people who don't hit the throttling. Of course as more and more high-bandwidth applications gain in popularity (blinkbox for instance) that throttling limit may become more of an issue.
Title says it all really.
It seems to be an endemic and almost unique feature of the telecoms industry (whether it be mobile phones, ADSL, or whatever) that the suppliers of such services just make stuff up (lie) as they go along, and seem to get away with it year after year. And when they're not lying then ...oh hang-on, that's an oxymoron.
Maybe someone should do a poll which asks the simple question "do you think you are being lied to and/or ripped off by your telecoms provider?". I wonder what the hit-rate would be on that?
I'm confused by how much it will actually cost me, not what I can get.
With everything tied in, meaning that I have to switch my phone service, and the only price quoted is only for the first 2 months at "only £10 a month", but the contract is for 2 years....
* actual speeds may vary depending on the quality of your phone line.
Consumers, try paying your ISP 'up to' whatever its monthly charge is and see how confusing and frustrating they find the practice.
How odd, there's no mention of their "up to 8M" Virgin Media National Broadband in the press release or on the web site.
That would be interesting, because it sucks.
When the term "up to" is used then I think many would agree there is an implication that one should be able to expect near to that figure most of the time, and to hit it on occasion. Otherwise the figure is meaningless. To hide behind the technically correct, "it only means, no more than", is as near to fraud as one can expect to get away with.
And again with claiming "unlimited" and then hiding a load of limits in the small print. That is clear fraud too, to any right thinking person. If the claim is unlimited then there are no deliberately applied limits. If deliberate limits have been applied then the claim can not be honestly made.
These are just blatant examples of how little the individual citizen is valued by society and those making the rules.
>I think many would agree there is an implication that one should be able to expect near to that figure most of the time
That's only one meaning of the phrase though. The English language also allows you to say "Vehicles powered by internal combustion engines can achieve speeds up to 400mph" - does that mean you should be annoyed that every car you've owned has struggled to get beyond 100mph?
I agree about the limits. I do think the adverts should be changed just not to an average speed. The advert should tell people to do a line check and see what they should expect, as you can't say everyone is going to get near those speed. I have to admit when I see up to I think up to not that I will be getting that speed but thats just me everyone sees things differently
The best answer would be to ban lock-in contracts (and all other techniques for making it hard to change ISP).
You sign up. You try the service. If it proves inadequate compared to the expectations created by the advertising, you cancel. No penalty, no delaying tactics, ho hassle.
Within a year the dud service providers will be going out of business, and the rest will be upgrading as fast as they can to make their reality resemble their claims before someone else nicks all their customers.
It depends where the speed drop is coming from. If it's connection speed then for xDSL changing provider won't help. The limiting factor there is your telephone line and assuming we're comparing equivalent technologies (ie;the same type of DSL) there'll be little or no difference.
Changing ISP only helps if the problem is poor throughput. Ideally you should get 85% of your connection speed as usable throughput. If you do get that when you want it then changing ISP is pointless unless it's to save money.
What would be interesting would be to see some charts (% subscribers vs speed) so people can see the distribution of these speeds rather than a one off value that does not really say how likely to are to actually get that 'typical speed'.
How about you force them to advertise nothing but the "Minimum speed they can garantee", so that if you get lower then that, you get to sue for breach of contract?
But then again, forcing them to do that would make sense so ...
The only ISP in the UK that owns the cabling is Virgin. The rest don't so it's not reasonable to expect them to be able to accurately predict its characteristics.
Getting an ISP to guarantee a speed would be like getting a taxi service to guarantee a journey time. Virtually impossible and certainly they couldn't do it in their adverts. Can you imagine it? "Use Bob's taxis - we get you to your destination in fifteen minutes. "Hello? I'd like to book a taxi to take me from Plymouth to Heathrow".
What we have here is the ultimate combination, ISP's are mostly a bunch of crooks and end users are mostly a bunch of kneejerk idiots.
Old dial up with Tiscali at 50 kb per second was mega fast at the time.
And we actually got 50 kb a second! =]
Ah, the good old days when you could leave your door unlocked and noone would steal your crap because they already had better at home...
doesn't mean my average speed for a journey on it is 60 mph
And this will help, how?
Last time I changed ISP, the website told me I'd get 9Mb/s - I queried this over the 'phone as I was getting around 2.7Mb/s and was told again, yes it should be around 9M/s, maybe a little less but certainly close to 9Mb.
Signed up... and got 2Mb/s.
So, exactly what is to stop the ISPs giving false estimates?
By the time your line has finished it's 'training period', it's too late to cancel.
nigel11 has it right - no tie-in period & the bad ISPs will disappear