Average UK household gets half the advertised broadband speed

[HEXUS]

Results come from 235,000 participants in a Which? consumer survey over the last year.

Read more.

[scaryjim]

the Which? results have probably been tainted by poor Wi-Fi connections in user dwellings.

Seems a reasonable shout to me. Whilst I currently get similar speeds on both 2.4GHz and 5GHz wifi, in the evening the 5GHz can be as much as twice as fast as the 2.4GHz based on speedtest results to the same server.

I honestly never thought I'd see the day that localised wifi interference would bottleneck my internet connection, but here we are...!

[DanceswithUnix]

I honestly never thought I'd see the day that localised wifi interference would bottleneck my internet connection, but here we are...!

My DMZ is on a 100Mbps switch, if broadband gets much faster I will have to replace it. Doesn't get much more first world problems than that

[Ozaron]

Damn, should have participated.

[Hoonigan]

I get above my quoted speed.

Paying for 350/20, receive 390/21.

[Xlucine]

Furthermore, the average person in a family might have their speedtest result slowed down due to other people using the broadband connection at the same time.

Does speed get evenly distributed between users on a typical home router? Because it's hard to appreciably suck up the majority of a 200 Mb/s connection without running a speedtest (or reinstalling your entire steam library, steam servers are a good bandwidth source), so you'd need the router to be failing somehow for the speed to be significantly affected. YT streaming at 1080p/60fps is only a few Mb/s (~20Mb/s with a ~33% duty cycle per windows task manager), you should be able to support dozens of simultaneous streams on a top-end connection

[Ozaron]

..(or reinstalling your entire steam library, steam servers are a good bandwidth source)

As someone with extremely limited broadband, Steam is NOT bad on this. Steam yields bandwidth on my network rather effectively, albeit with me being the only user, especially to services like Youtube or Twitch or what have you that require it on demand. Battlenet, or Origin, or even Uplay, ALL refuse to yield bandwidth. If I download with any of those clients and try to reach Google, the connection simply times out, and Youtube / Twitch etc are impossibilities.

If you're just saying Steam's pretty good at maxing out a fast connection, I have no experience ( ), but people bash Steam network more often than it deserves IMO.

[Tabbykatze]

I get above my rated speed but all statistics garnered from a client on WiFi should be omitted.

I had a Customer saying he'll get the BT Infinity 200mbps package and that he is going to be quoted an average of 100mbps. I guffawed and said I would only pay BT for the 100mbps package then because that's BS. People let providers get away with so much sometimes!

[Corky34]

Something does seem fishy with Which's results, IDK for certain as i can't find any full results but i just ran their speed check and compared mine to others in my area for Virgin and every comparison showed an advertised speed of 200Mps, are they ignoring the speed people are paying for and just using the maximum advertised speed of each provider?

[DanceswithUnix]

Does speed get evenly distributed between users on a typical home router? Because it's hard to appreciably suck up the majority of a 200 Mb/s connection without running a speedtest (or reinstalling your entire steam library, steam servers are a good bandwidth source), so you'd need the router to be failing somehow for the speed to be significantly affected. YT streaming at 1080p/60fps is only a few Mb/s (~20Mb/s with a ~33% duty cycle per windows task manager), you should be able to support dozens of simultaneous streams on a top-end connection

No, because Quality of Service is hard to do specially when these days most clients are mobile devices on WiFi at varying distance from the router or AP, and web pages are designed to pull a gazillion lumps of information from multiple sources across multiple connections which is like a denial of service attack on your router.

[MrRockliffe]

Does this take into account using WiFi etc? Yes, I'm on Virgin 200 down, and get immense speeds all the time.

http://www.speedtest.net/my-result/d/48290794

That's over WiFi - over ethernet I get a bit more. However, if I'm in the room furthest away from the main router, I get more like 170 down (using a mesh network).

[satrow]

Something does seem fishy with Which's results, IDK for certain as i can't find any full results but i just ran their speed check and compared mine to others in my area for Virgin and every comparison showed an advertised speed of 200Mps, are they ignoring the speed people are paying for and just using the maximum advertised speed of each provider?

Judging by the BT speeds given in my locale, Which seems to be aware of the difference on VDSL, maybe not for Virgin Cable though.

Provider/Distance(km)/Advertised/speed(Mbps)/Download speed(Mbps)
Vodafone 0.3 76.0 31.3
Vodafone 0.3 76.0 31.1
Vodafone 0.4 76.0 17.3
BT 0.4 76.0 16.1
BT 0.4 50.0 11.4
BT 0.4 76.0 10.7
BT 0.6 50.0 53.1
BT 1.0 50.0 20.3

Guess which of those results are probably connected somewhere along the 500+ meters of aluminium cable...

[Iota]

Something else aside from wifi may well be affecting those results, I was having some issues with both speed and my line speed being reduced. Eventually narrowed it down to my powerline adapters triggering the line management to try to provide a "stable" speed, which was almost a half of my advertised speed. It kept dropping it down to the minimum allowed on the line.

After four engineer visits, we finally found what was causing the problem, since getting rid of them I get my advertised speed even during peak times.

[peterb]

But what are they measuring? Line speed to the exchange? Speed to a given server, or an average spped to several servers? And under what conditions? There are just too many variables and an uncertain methodology - at least without reading the original article.

[bledd]

WiFi and powerline (and some stock routers) should not be included in these tests. But it's impossible to filter them out.

I've always had max expected speed (with a couple of percent) at every property, even ones that are 5km from exchange.

Keep modem(router) right next to the phone port and use a 0.5m Adslnation rj11 cable = win.

[Xlucine]

Stock router should always be included - if the broadband company doesn't provide a router capable of providing their advertised speeds, then that's their own problem