Broadband News

Is it a hollow victory with the broadband speeds in advertising win!

Broadband advertising we are being told is now more accurate, but from our viewpoint while we have seen average peak time speeds adopted by the major providers the massive long tail of smaller providers continues with the old up to figures.

For some years prior to May 2018, broadband adverts if they showed a service speed had to state the speed that 10% of customers on the service could achieve, the change in May was two fold (1) the speed must be the median average (2) based on what people can get during the 8pm to 10pm window. A common claim is that this is now more accurate but the same number of people are getting the 50% speed as were getting the 10% speed, what the change means is an additional 40% should see speeds above that shown in the adverts. The adverts now are no more accurate at telling an individual what speed they will get. The important part seems to be though that people can feel as if something has changed for the better.

The key to learning what to expect is only available once you have shared your address and/or telephone number with a broadband provider as part of the sign-up process or during an order enquiry. This process is governed by the Ofcom Broadband Speeds Code of Practice and is often referring to connection speeds rather than peak time throughput speeds of the adverts.

We publish a longer table of service providers speeds at the start of each month, and later in the month the average speeds we are seeing at peak times. The speeds we report for the packages is lower than what the providers use in their adverts and while Wi-Fi and broadband problems account for some of this and using a limited number of lines to derive a UK average can mask an evening or two in a month or quarter i.e. if in a 90 day period a line was slow for just 3 evenings the median average for that line is not going to be moved and in this scenario the actual customer will most likely remember three evenings when video streaming and gaming were impossible and we also tend to see more people when they are having a problem.

Broadband Speeds for the most popular UK broadband packages during the 8pm to 10pm period in July 2018
Ordered by median download speed
All speeds are in Mbps - Mega bits per second
ProviderQualityDirectionSpeed of bottom 10%Speed of bottom 20%MedianMeanSpeed of Top 20%Speed of Top 10%
Virgin Media VIVID 200A (0.6)Down 69.4 86.4 176.7 151.2 205.2 215.1
    Up 11.5 11.9 15.7 16.2 21 21.4
Virgin Media VIVID 150B (0.9)Down 65.2 72 100.5 104.5 136.1 153.7
    Up 8.2 9.6 11.6 11.7 12.4 13.3
BT Superfast Fibre 2 (FTTC 80/20)A (0.2)Down 35.4 42.3 55.2 54.3 68.5 71.5
Up 10 11.3 16.4 15.1 18.3 18.6
Virgin Media VIVID 100A (0.6)Down 27.9 35.9 52.2 62.1 100.7 107.4
    Up 5.5 5.7 6 6.9 8.7 10.9
Plusnet Fibre Extra (FTTC 80/20)A (0.2)Down 33 38.5 49.4 50.4 63.7 70.9
Up 9 9.7 13.4 13.6 18 18.2
Vodafone Superfast 2 (FTTC 80/20)A (0.2)Down 29.5 38.7 47.2 49.5 63.1 72.7
    Up 8.5 9.9 14.8 13.8 17.5 17.8
Sky Fibre Pro (FTTC 80/20)A (0.4)Down 33.3 39.7 48.3 49.1 59.7 69.6
Up 7.4 9.4 15.4 14.1 18.2 18.5
TalkTalk Fibre Plus (FTTC 80/20)A (0.3)Down 28 37.7 45.7 47.2 62.1 68.3
    Up 7.2 9.5 15.4 14.3 18.2 18.5
Virgin Media 70B (1.1)Down 15.2 17.8 32.6 37.1 56.8 67.6
    Up 3.6 3.9 4.7 4.6 5.2 5.4
EE Fibre (FTTC 40/10)C (1.5)Down 9.7 15.7 26.9 24.7 33.8 35.6
    Up 2.2 3.2 5.2 5.3 7.6 8.2
BT Superfast Fibre (FTTC 55/10)A (0.5)Down 11.4 15.9 26.2 26.9 37.9 44.4
    Up 2.2 3.2 5.1 5.3 7.6 8.7
Vodafone Superfast 1 (FTTC 40/10)A (0.5)Down 7.8 14.1 24.5 23.4 33.5 35.5
    Up 2.2 3.5 5.4 5.4 7.6 8.1
Plusnet Fibre (FTTC 40/10)A (0.5)Down 10.9 14.9 24.2 23.8 33 35.7
    Up 1.9 2.7 5.1 5.1 7.4 8.3
Sky Fibre (FTTC 40/10)A (0.5)Down 9.6 14.2 24.2 23.3 32.7 35.6
    Up 2.5 3.4 5.6 5.5 7.8 8.4
Virgin Media 50B (1.1)Down 2.7 5 19.2 23.6 46.3 53.3
    Up 0.7 1.2 2.3 2.1 3 3.1
EE ADSL/ADSL2+B (0.8)Down 1.1 2.2 6.7 7 11.3 14.9
    Up 0.3 0.4 0.7 0.6 0.9 0.9
Sky ADSL2+A (0.7)Down 1 2 5.6 6.8 11.8 14.6
    Up 0.2 0.4 0.6 0.6 0.8 0.9
Plusnet ADSL/ADSL2+A (0.6)Down 0.6 1.6 5.6 7.2 13.9 16.2
Up 0.3 0.3 0.6 0.6 0.9 0.9
TalkTalk ADSL2+B (0.8)Down 0.7 1.5 4.9 6.1 10.6 13.7
    Up 0.2 0.3 0.6 0.6 0.7 0.8
BT ADSL/ADSL2+A (0.8)Down 0.6 1.5 4.4 6 10.3 14.2
    Up 0.2 0.3 0.5 0.5 0.8 0.9

We apologise for adding more columns of figures with the speeds of the bottom and top 20% now showing, but we feel that the additional columns will help to show the spread of speeds we are recording and an observation to make is that invariable the advertised speeds see to be close to what we see for the top 20% figure, so tracking this over time may prove worthwhile.

The table is of course shorter than our full monthly summary published at the start of the month, since filtering for the two hour period between 8pm to 10pm means there is a lot less data available, this is why none of the FTTP services show up, the BT Superfast 2 FTTP service does occassionally just reach the minimum number of tests to be included.

The quality column needs its usual small explanation, to get a grade A score a speed test needs to very quickly reach the average speed and once at that speed sustain it for the period of the test. Since we are testing from a device to the persons broadband router and across the connection and Internet to our server we also include factors like Wi-Fi performance and how good a providers internal network and peering links are. A grade A test looks like this and at the end of the scale (beyond grade E in this case) there are tests like this, a good example of a Grade B test is this one. The way the grading works means it is looking at both throughput and the time it takes for a connection to start transferring data e.g. slow DNS look ups that delay the start of the download will register, while many speed testers ignore this delay even though it would have a massive impact on activities such as web browsing. At the end of our speed test an analysis page in the test will tell you your quality score and also how your line compares with others using the same technology.

Regular readers and the observant will notice that the FTTC services from EE often show a lot lower quality score than other services and this is because of something unique in how the service seems to behave, various examples are Test A, Test B, Test C, Test D which show a controlled rise in the speeds generally hitting the ideal speed around 3 seconds into the test. The speed test is fetching HTTP/HTTPS based data and if there is a slow rise to the top speed people will notice this as larger images on websites are slow to load. There are some EE customers still on FTTC that are not affected (e.g. this test) why some are not affected is a big unknown, it does not seem to be a Wi-Fi based issue.

Comments

Are the column titles for the top 10% and top 20% correct?

  • Michael_Chare
  • 11 days ago

They are now.

  • andrew
  • thinkbroadband staff
  • 11 days ago

Thank you. I am pleased to see that my speed test results for Vodafone put me in the top 10% despite the fact that I do them using wifi.

  • Michael_Chare
  • 10 days ago

Post a comment

Login Register