Skip Navigation


Advertised speed versus actual speed
Thursday 22 November 2007 19:50:26 by Andrew Ferguson

We seem to enjoy being second or lower in the UK. It acts as a chance for people to vent their anger about the state of UK broadband and how we get measly speeds compared to the rest of the world. The question is are we comparing the same two things?

When you look at data from places like The Information and Technology Innovation Foundation (published in April 2007), we do badly coming 17th out of 30. People wish for the 100Mbps connections or sometimes 1Gbps connections of Japan and Korea, but what we forget is that the average speeds quoted are usually the advertised speeds. So it is likely that most UK consumers are reading data like this and comparing the actual download speeds they get versus what is the headline advertised rate in the other country. One example is a writer for The Times who moans of only 5Mbps from a 20Mbps connection in Italy.

If lengthy studies were to be done to look at actual experienced download speeds, would the UK be a lot further up the table? Probably not, but the reality that 100Mbps connections probably rarely meet those speeds when downloading from web sites would be more apparent. Consider this- if a country has one million people connected at 100Mbps, how many web sites have the connectivity to support even 10% of these people downloading at the same time?

UK xDSL providers already have the systems in place that allow them to provide a reasonable estimate of possible speeds by accessing line data provided by Openreach. The public can access a version of this at www.btwholesale.com/getbroadband, but this tells you nothing about how a provider will perform on Friday nights when you probably use your connection.

Sites like ourselves go some way to helping people share their experiences and discuss actual speeds for different internet applications, but how can this be expressed in advertising in a way that is consistent across all providers? Perhaps you can't and what is needed is education to make more people aware that the advertised speed is a theoretical maximum with many factors affecting your actual experience. Companies like Epitiro try to make thousands of measurements to compare providers and has recently won a contract to do this for the regulator in New Zealand, but again the data, while useful, often does not directly translate to the experience of the individual broadband user.

With the increasing roll-out of ADSL2+ and potentially other xDSL variants in the future, consumers will need more help to figure out whether what looks like a three times faster package would be better for them or not. To this end, one example would be to publish the connection speed profile for a provider. Entanet does something like this already by publishing a chart showing the IP Profiles its customers currently have. From this you can deduce that 50% of people connect at speeds of 4544Kbps or higher with 21.5% hitting the absolute maximum connection speed of 8128Kbps.

At the end of the day, the UK broadband market is a competitive one with providers competing for our business. This competition can push advertisers to try and find data to prove their point. Regulating the advertising is difficult and as with any rules, companies will look for ways to exploit or bend the rules to meet their needs. The biggest force is us the consumer. If we are not happy, let friends know and at the first opportunity change provider to one that suits your needs better.

Comments

Posted by hairyman over 9 years ago
I think we are missing a little more than connection speeds, how about actual thoughputs averaged and compared over a range of times of day, size of download and source of download are important.Maybe the New Zealand comparision arrangement is the way to go!!
I connect at 8128k but rarely get more than 1meg probably due to exchange contention?
Posted by chrysalis over 9 years ago
in korea and the like tho the actual connection speed is whats advertised. Service contention is something that we should accept happens worldwide and long distance downloading will be slower.
Posted by herdwick over 9 years ago
"in korea and the like tho the actual connection speed is whats advertised." same here - 8128. If only ever getting 1Mbits/s of throughput I would be looking to change ISP.

Korea and Japan are essentially giant LANs, their cultural and language / character sets are somewhat unique and therefore the bulk of the traffic is internal.

Japan's international transit capacity is not very impressive - we have at least 5 times more.
Posted by hairyman over 9 years ago
Hi herdwick. I have tried changing isps during the evening it makes little difference who I have used, but in the daytime in what I guess is non peak rate on our exchange I get 4 to 5 meg. I am now using one of the top five rated ISP as rated at DSL and here. Its all down to the exchange and the contention. It only has BT / openreach gear installed. Had broadband for over 4yrs now it has got slower since it became popular about two years ago.
The UK average "real speed" has been estimated as a around 1.3mbps. Dont forget we still share our connection with up to 50 others.
Posted by Gypsydog over 9 years ago
Re:"[i] to make more people aware that the advertised speed is a theoretical maximum [/i]"
Perhaps Ofcom and the ASA should INSIST that the 'theoretical maximum' - allied to a codicil stating that you WILL NOT get this figure - should be the ONLY statement relating to 'speed' that is permitted.

Posted by CARPETBURN over 9 years ago
quote"The UK average "real speed" has been estimated as a around 1.3mbps"
Dunno if that is true or not but reading the story and assuming the enta situation is roughly the same for most ISPs it does seem the average is around 2-3Mb which to be honest is pathetic.
Posted by Going_Digital over 9 years ago
The actual performance can often have a lot to do with the end users equipment. It is not unusual to find quite a few home installed extensions using cheap alarm wire rather than a proper twisted pair cable. In a way it is a shame that the engineer install including a faceplate is not the standard install any more as it can make a huge difference. I would encourage everyone to test their connection using the test socket in the back of the NTE-5 master to see what speeds you really can get.
Posted by carrot63 over 9 years ago
Providers should be made to state average connection speed obtained by customers, rather than a theoretical maximum. Certain providers constantly deliver poorly, yet 'up to' allows them to claim otherwise.

"with providers competing for our business"

The competition ceases the moment you sign up to an increasingly long contract, after which the reality of terrible service sets in, and they are free to wipe their boots on you for the duration. Genuine competition should be quality as well as price based, but any difference is currently obscured by dishonest advertising.
Posted by CARPETBURN over 9 years ago
quote"The actual performance can often have a lot to do with the end users equipment. It is not unusual to find quite a few home installed extensions using cheap alarm wire"
Stop right there, my wiring is fine its the ISPs that throttle and cap and the old rubbish BT copper network thats the problem for most people. To suggest the whole country has faulty home wiring and thats why the average speed is around 2-3Mb is laughable.
Posted by ph00ey over 9 years ago
Go to the BT - get broadband link, enter your number and see what you can get (indicative) then click what next...


Active Server Pages error 'ASP 0131'

Disallowed Parent Path

/broadband1/checker/decision.asp, line 4

The Include file '../includes/adovbs.asp' cannot contain '..' to indicate the parent directory.

Someone's turned off parent paths in IIS and not told anyone to regression test their code. What a bunch of muppets
Posted by dayday01 over 9 years ago
I have had speeds of under 400kbps on my"up to 8mb"line for over a month. After two 10-day "stabilisation periods"my IP Profile is still limited to 250Kbps. I don't know what else can be done - my ISP seems to be powerless in the face of BT's inability to raise my speeds. Any helpful suggestions?
Posted by Jamiepw over 9 years ago
When the length of the cable from the exchange is mentioned, most people can assume that there may be a few joints between the two locations, a few years ago i had a problem on my line and when i eventually argued the case with BT that is was not internal to my house they eventually sent out an engineer.

He found that there was at least four locations where there may be a fault due to these locations having joints in them, it turns out that the connecting strip crumbled in his hand when he examined the joint.

Is it any wonder that we have poor speeds with this type in infrastructure?
Posted by Jamiepw over 9 years ago
Time and time again the ISP's mention the headline grabbing figures and not the real average figures, its all Spin Spin Spin.......

Wireless is the way to go!!
Posted by terryl over 9 years ago
If BT spent money in this country on their infastructure, instead of piss balling about blowing the cash on shakey ventures around the world things may be very different. More Exchanges thus shortening line lengths would help the situation, but so would using optical rather than copper. It's about time the advertised speed rates were clobbered a bit more after all not everybody can live 10 meters from their exchange.

That's my winge over with
rgds to all.
Posted by andrew (Favicon staff member) over 9 years ago
For those that forget we did an average speedtest earlier in 2007 and it came out at 2.3Mbps. Now this could be called pathetic by some, but given the distance factors of ADSL seems reasonabe. What sort of figure do other countries using ADSL get?

Finding this sort of info is difficult, we could try and look at the average speed for all overseas speedtests that are done here perhaps.
Posted by herdwick over 9 years ago
"Dont forget we still share our connection with up to 50 others." that was never true and it certainly isn't now. 50 of you might share capacity with up to 2500 was perhaps a better analogy although these days you each have less than 50 kbits/s of bandwidth at the ISP end so the ratio is more like 80:1 with a 4M connection.
Posted by hairyman over 9 years ago
Ref the poor quailty of internal wiring , I personally wouldnt recommend anyone to connect a broadband modem/router to anything other than the mastersocket. In these days of wifi or wired domestic networks this makes sense.There would need to be serious faults or very long internal cables to alter the connection speeds at the relatively low frequencies used by bb.

Posted by hairyman over 9 years ago
There is no magic hapenning by using twisted pairs over cables such as what I call "figure of eight" , both can provide low loss and low interference rejection if terminated correctly and use copper. Interconnections are the places or loss as are high resistance or poor joints that can rectify signals passing through them. Coaxial or optical cable would be the good option.
Posted by tcrooks3843 over 9 years ago
The task of overhauling the infrastructure is galactic. By and large BT are doing an adequate job but they would do a lot better if Openreach wasn't a 'closed society' that the consumer has difficulty in reaching and influencing.

BT engineering probably needs to be restructured so that more attention is given to local network issues over backhaul ones. The balance seems just wrong. I base this on the frequency of cries for help with problems that stem from the bit of the network from exchange to master socket.
Posted by keith_thfc over 9 years ago
Re Speeds - Virgin claim to be the "UK's fastest broadband"

http://allyours.virginmedia.com/websales/service.do?id=2

How can this be? I though their max speed offered was 20 MB which is lower than that of be*.
Posted by andrew (Favicon staff member) over 9 years ago
I am led to believe this is based on Epitiro information and most likely an average, rather than outright fastest connection.
Posted by CARPETBURN over 9 years ago
quote" am led to believe this is based on Epitiro information and most likely an average, rather than outright fastest connection."

Or in other words its technically another virgin media lie and mis-advertising.
Posted by skyguy over 9 years ago
How about TB rating ISPs by evening download speed as a percentage of client max speed? Can the speedtest determine the BT Profile for the line? That way the ISP doesn't get marked down by clients on slow lines.
Posted by wlchubb over 9 years ago
There is a huge difference between the profile of your line and actual download. As I understand it the profile is the theoretical maximum capacity. In my case, and that of the majority of my neighbours, the downstream quoted by the BT Home Hub is pretty constant at 2,528 Kbps whereas the actual varies between 250Kbps and 1.8Mbps. I had BT investigate recently and they explained, off the record, that it's all down to the dramatic uptake of BB from the same BT exchange combined with the limitation of the old copper cable. I.E Contention.
You must be logged in to post comments. Click here to login.