A new report has used data from 122,845 consumer speedtests to identify both the top 10 fastest and slowest UK streets for broadband download speed, which claims that Kingsclere in Huntington (York) was the “slowest” on 0.22Mbps and Darwin St in Livingston (West Lothian) was the “fastest” at 182.52Mbps. Beware caveats.
The uSwitch.com research is based on speed tests conducted between 1st October 2018 to 30th September 2019. In order for a street to qualify for inclusion, tests from at least 3 unique IP addresses and at least 10 residential properties were required at a postcode.
Overall the study claims that 19% of the data sample struggled with speeds of less than 10Mbps (down from 26.3% last year), while 9% crawl along at less than 5Mbps (down from 13.3%), but happily the number of broadband users enjoying faster speeds is growing. Nearly a third of users (38%) now get “superfast broadband” speeds of 30Mbps+ (up from 31% last year and 22% four years ago).
As usual one of the biggest problems with studies like this is that they can easily be taken at face value, which is a problem because consumer speedtests don’t correctly reflect the actual underlying availability of faster networks.. and uSwitch agrees.
At present around 96% of the UK can access a so-called “superfast” service and it’s similarly noted by uSwitch that such networks were actually available on 7 out of the 10 slowest streets below (70%). Despite this a recent survey by the same site found that only 61% of British people thought they could access such a connection. Clearly awareness is still a key problem.
In other cases consumers may be aware that a faster service exists but they have simply chosen not to upgrade due to various issues, such as the potentially higher price, being stuck in a long 18-24 month contract term or a simple lack of need / desire for anything faster than what they have today.
Speedtests can of course also be influenced by other factors, such as poor home wiring, user choice of package (e.g. 1Gbps could be available but most people may still pick a slower and cheaper tier on the same fibre), local (home) network congestion and slow WiFi performance etc. In short, take these results with a good pinch of salt.
The Top 10 Slowest UK Streets for Broadband
Rank | Street Name and Location | Average download (Mbps) | Superfast broadband available? |
1 | Kingsclere, Huntington, York | 0.22 | Yes |
2 | Monksfield Way, Slough | 0.22 | Yes |
3 | Ash Lane, Whitchurch | 0.32 | To some |
4 | Dunlop, Kilmarnock | 0.33 | No |
5 | Canisbay, Wick, Island of Stroma | 0.39 | To some |
6 | Duiletter, Colintraive, Argyllshire | 0.46 | No |
7 | Eland Way, Cambridge | 0.47 | Yes |
8 | Ansley, Nuneaton | 0.49 | To some |
9 | Quarterland Road, Killinchy, Newtownards | 0.49 | No |
10 | Malmesbury Park, Runcorn | 0.54 | Yes |
The Top 10 Fastest UK Streets for Broadband
Rank | Street Name and Location | Average download (Mbps) |
1 | Darwin St, Livingston | 182.52 |
2 | York Road, Elvington, York | 135.56 |
3 | Chestnut Grove, New Malden | 134.34 |
4 | Mallards Road, Woodford Green | 110.17 |
5 | Albemarle Road, St Ives | 103.94 |
6 | Barnsley Road, Woolley, Wakefield | 92.05 |
7 | Central Park Road, East Ham, London | 86.22 |
8 | Oakhill Road, Horsham | 82.41 |
9 | Baberton Mains Drive, Edinburgh | 76.25 |
10 | College Gate, Cheltenham | 76.09 |
We should add that business class high-capacity leased lines might be impacting some of the above the results and it’s similarly unclear whether mobile connections were included or not. Lest we forget that the slowest streets will be those precious few isolated locations that still struggle to get a working internet connection at all or have too few results, which won’t show up above.
Dunlop has FTTC, but lots of long lines.
Taken with a HUGE pinch of salt, considering Kingsclere is FTTP enabled then the reason for it being so slow is either users not upgrading to a more advanced network or they have some pretty poor wiring somewhere. To be honest I am surprised the list is not dominated by FTTP locations or is this ‘sample’ only covering those locations still connected to Copper/Hybrid networks instead of Pure Fibre?
I notice the requirement for 10 location tests in a postcode will exclude many locations, particularly rural.
I think it means there must be 10 properties in a postcode.
Here in Devon we have some postcodes with only one property. My own postcode has only 9 residential and 3 business premises. Hardly a fair test.
My own postcode has two properties on it.
Ah, the caveats killed it.
These results don’t take into account the Kingston -upon-Hull area where the average speed of broadband via lightstream is in excess of 200 mbps with other packages offering up to 400 mbps