The Best and Worst 10 Areas in the UK to Live by Broadband Speed
A new study has identified the top 10 and bottom 10 council areas in the United Kingdom to live by their average (mean) broadband ISP speed, which is ranked by considering the average speed that is possible if everyone in the area brought the fastest available service. Speedtest data is then used as a balance.
In the past most studies of this sort have tended to rank areas by looking at the results from speedtest based data and then producing a direct average, which can be highly misleading because such reports often fail to factor in the underlying network availability (i.e. whether or not local users could order a much faster connection).
Alternatively other studies have attempted to rank areas purely by network availability, which doesn’t always reflect the fact that in some areas even the best available networks may not always deliver its top broadband speeds. By comparison this research from Thinkbroadband has focused on network availability but also weighted that against the known speedtest based performance in order to produce a potentially more useful table.
One difficulty here is shown by the fact that the central City of London area pops up mid-table in the fastest 10 list, which is unusual given that it has a very low coverage of “superfast broadband” (30Mbps+). “This highlights the effect that over 1 in 3 premises with access to [FTTP] with maximum speeds ranging from 300Mbps to 940Mbps can have,” said TBB’s Andrew Ferguson. The actual observed speeds for the area thus tell a very different story (right side of the table).
NOTE: We’ve had to cut the ultrafast (100Mbps+) and FTTP/H coverage columns out as otherwise the data wouldn’t fit (TBB has the full table).
Top 10 UK Council Areas for Best Case Speeds
| Council | Best Case Mean Average Download (Mbps) | Superfast Coverage 30Mbps+ | Observed Median Download (Mbps) | Observed Median Upload (Mbps) | Observed Mean Download (Mbps) | Observed Mean Upload (Mbps) |
| Bournemouth | 401 | 99.70% | 25 | 4.7 | 37.3 | 6 |
| Tower Hamlets | 359 | 93.10% | 14.6 | 2.3 | 32.3 | 16 |
| Wandsworth | 358 | 98.20% | 22.7 | 4.6 | 42.1 | 20.3 |
| West Berkshire | 355 | 98.40% | 25.7 | 5.7 | 37.4 | 14.3 |
| City of London | 343 | 51.60% | 19.1 | 6.5 | 32.9 | 23.2 |
| Newham | 321 | 98.50% | 23.4 | 4.9 | 39.8 | 13.5 |
| York | 321 | 95.60% | 21.3 | 4.5 | 36.5 | 8.3 |
| Southwark | 319 | 93.70% | 17.9 | 2.6 | 33.4 | 11.2 |
| Greenwich | 312 | 99.10% | 25.3 | 5.4 | 37.3 | 11.9 |
| Hounslow | 311 | 99.10% | 25.7 | 4.8 | 40 | 10.1 |
Bottom 10 UK Council Areas for Best Case Speeds
| Council | Best Case Mean Average Download (Mbps) | Superfast Coverage 30Mbps+ | Observed Median Download (Mbps) | Observed Median Upload (Mbps) | Observed Mean Download (Mbps) | Observed Mean Upload (Mbps) |
| Moray | 40 | 85.80% | 15.4 | 1.8 | 19.4 | 3.6 |
| Dumfries and Galloway | 40 | 82.50% | 14.1 | 1.1 | 18.7 | 3.7 |
| North Norfolk | 40 | 84.10% | 14.8 | 1.3 | 20.2 | 4.4 |
| Argyll and Bute | 40 | 80.30% | 13.1 | 1.6 | 18.7 | 3.9 |
| Highland | 39 | 78.60% | 13.6 | 1.5 | 19.6 | 4.1 |
| Aberdeenshire | 39 | 80.60% | 17.2 | 3.6 | 22 | 5.2 |
| Na h-Eileanan an Iar | 38 | 70.40% | 15.6 | 1.8 | 20.2 | 4.3 |
| Fermanagh and Omagh | 35 | 63.90% | 11.8 | 1.2 | 19.3 | 4 |
| Shetland Islands | 35 | 73.50% | 13.9 | 0.9 | 18.7 | 4 |
| Orkney Islands | 33 | 65.40% | 8.3 | 15.3 | 0.7 | 2.7 |
Mark is a professional technology writer, IT consultant and computer engineer from Dorset (England), he also founded ISPreview in 1999 and enjoys analysing the latest telecoms and broadband developments. Find me on X (Twitter), Mastodon, Facebook and Linkedin.
« Snooping TalkTalk Security Feature Causes Customer Concern UPDATE2
Latest UK ISP News
- FTTP (5512)
- BT (3514)
- Politics (2535)
- Openreach (2297)
- Business (2261)
- Building Digital UK (2243)
- FTTC (2043)
- Mobile Broadband (1972)
- Statistics (1788)
- 4G (1663)
- Virgin Media (1619)
- Ofcom Regulation (1460)
- Fibre Optic (1394)
- Wireless Internet (1389)
- FTTH (1381)
If I read it correctly,it includes business and residential premises and so can be so heavily skewed that it seems a pointless set of figures. A handful of Gigabit connections is going to send an area up the tables whilst really it should be a lot lower going on purely residential connections.
Any skew is usually visible in the mean being massively different to a median speed.
NOTE: Businesses with a leased line are not included in the best case calculation, since they are on demand type connectivity.
For the top 10 its the presence of native full fibre e.g. York TalkTalk UFO, Bournemouth Gigler and the list goes on.
This article is a bit incorrect in that many parts of London still do not have fiber optic lines and have no choice but to still use ADSL.
It’s reflected in the relatively low superfast and median download in the table. Not super surprising as much of it is commercial space and they tend to just need a connection, not a good connection.
Most of London doesn’t have fibre-optic lines, but only DSL or cable, with a few exceptions where some alternative providers like CityFibre have built their own fibre networks.
Sadly misleading, as BT have left the acceptable speed ranges for FTTC so wide.
ISPs cannot even raise underperforming lines as faults because according to BT the speeds are within “expected limits”.
This is a con on a massive scale.
Absolutely nothing to do with acceptable ranges speed ranges Openreach quote to BT Wholesale.
No con intended, intention to show that many people are not buying the fastest services available to them and if everyone did what the potential impact on the overall results are.
You may have misunderstood, the “con” refers to FTTC itself, not your statistics.
If by “fastest services available to them”, you mean the actual speed attainable with each person’s specific line, then the statistics are meaningful.
If you refer to headline rates, then the statistics are not meaningful as many lines do not deliver the headline rate and BT have no intention of fixing them.
The stats take into account distance so vdsl2 has varying speeds from 0 to 76 based on distance
There is NO assumption that all vdsl2 delivers 76 Mbps
It will be good when the new rules come in and median download speeds are used, which should get rid of the upto 76Mb confusion and replace it with more accurate average speeds.
@Andrew I imagine those distance based speeds are headline rates…. I doubt if they are values from the lower end of the “impacted range” for a line. In which case – rather meaningless.
We need statistics where actual line performance is used. If a premise has never ordered FTTC then a middle of the range estimate could be used.
I guess only Openreach would have the “real” statistics…
Again
OUR Model for speed versus distance on fttc has had several iterations
It is based on looking at speed test results at the various distances to arrive at the model.
When checking against Openreach it usually lines up with low end of their range.
The Median figures of 15-25Mb (approx) seem about right to me around 30-40Mb is about the typical you see for FTTC unless you are in the top percentiles.
Unfortunately any summarises of this kind are only broad indicators. “if everyone in the area brought the fastest available service” itself dismisses the actual speed that people can achieve and that the given supplier is actually available at their premises. TBB could have adjusted the copper based products (FTTC and ADSL) as they could cross reference with speed test data or simply apply a formula.
As for BT Openreach they should now have sufficient data to determine where individual pairs or cable lengths are under performing in comparison with others on the same cabinet and investigate. Currently there appears to be no incentive for Openreach to be proactive. We all accept that these upto products are dependent on distance but perhaps they should be divided into speed bands at differential pricing with larger discounts than currently proposed by Vodafone. Whether this would be enough incentive for OR.
tbb did adjust the ADSL2+ and VDSL2 speeds, i.e. 17 Mbps and 38 Mbps are not assumed across the board, but estimated speeds are used, so you get 3, 4, 8, 12, 20 Mbps ADSL2+ lines all contributing.
Stand corrected.
Now this explains why I do not have Fibre supported since only 51.60% of City of London can achieve 30Mbps or more. While the rest of London have it at over 90%+.
Even with FTTC this speed is unachievable in my case simply because my building is located 800 meters away from the Bishopsgate exchange, but the Bishopsgate is an Exchange Line Only. Meaning the speeds will be max 28Mbps from 800 meters distance if it was theoretically possible.
So when I contacted BT Openreach via email they told me the so-called good news which I wasn’t really excited about. As quoted “However, the good news is that there is an ongoing project to divert your line through a traditional green cabinet, which will allow you to order fibre broadband once completed.”
Traditional Green Cabinet means that what BT Openreach will offer at best would be FTTC or G.Fast. If it is FTTP then a green cabinet would be avoided. Not to mention BT don’t mention exactly when works will start. What’s actually very disappointing is that on their site it shows “You’re in a plan to get Superfast fibre but we haven’t started work yet.” This has been that way for over 4 years now, so it’s no surprise that the speed stats are so low in the City of London. The only way I’ll achieve 30+Mbps never-mind 100+Mbps via FTTC would be a green cabinet located very close to the property block. But that is of-course not the solution because it will still affect those who live faraway from that cabinet.
There is also no timetable, plan or diary record on a calendar of work schedules. I know from personal experience when you don’t maintain a diary or record of your routine schedule then nothing ever gets done. This is the exact situation with BT. For example they mention that they’ll build 3 million premises with FTTP by 2020 in 8 cities but they don’t mention where exactly these works will take place and when these particular works will be made in the next 2 years.
So it appears that in order for the speed stats to improve in the City of London. BT will have to first achieve 100% superfast coverage in the rest of London before they begin to work in the City of London. And if BT don’t do the works properly and in time then our only rescue is alternative Fibre providers. The only real hope in jumping those speed stats.
@Rahul – would you really expect a company to publish ‘a timetable, plan or diary record on a calendar of work schedules’?
@TheFacts: No, not the calendar of their work schedules. But I would expect the dates and times of their work completions. For example when you see BT Service Status with a list of areas with the service having gone down. BT specify in brackets (estimated clear date….) when they’ll work to solve the problem and the dates and times of the resolutions once done.
We also see in Hyperoptic Registering Interest Going live (Blue) Installation agreed (Orange) Registering interest (Grey). Inside any of these properties from the map you’ll see for example “Please be aware this takes between 3-9 months.” With check-boxes of the 5 steps with tick boxes shown in each of them which gives the idea of when a project will be completed and thus go live.
We don’t see any of that with BT Openreach Fibre Journey Checker. All it says is that there is a plan (IN SCOPE) but they have not started any of the works yet and this is why there’s so much delay in their projects. BT Openreach gives us an impression of uncertainty because we don’t know whether they’ll really fulfil the projects they claim to set out or suddenly change their minds and abandon the project(s) altogether.
Well done Andrew for you and your team putting these figures together. However for jo public, MPs, Which readers, etc. they will find the figures difficult to understand and will not want to spend the time understanding the subtleties of median and mean and this is proved by some of the comments listed above. Jo public automatically thinks median and mean is roughly the same thing so could I suggest remove all references to mean figures as in most cases the reader will be thinking ‘the average’ and median better represents this.
For me I would like to see a report like this…
COUNCIL,
GEN3 COVERAGE, BEST CASE MEDIAN DOWNLOAD, BEST CASE MEDIAN UPLOAD,
GEN2 COVERAGE, GEN2 MEDIAN DOWNLOAD, GEN2 MEDIAN UPLOAD, GEN2 PERCENTAGE THAT COULD UPGRADE TO GEN3,
GEN1 COVERAGE, GEN1 MEDIAN DOWNLOAD, GEN1 MEDIAN UPLOAD, GEN1 PERCENTAGE THAT COULD UPGRADE TO GEN2.
And GEN1, GEN2 and GEN3 refer to…
BROADBAND GENERATION 3 = would cover G.fast, DOCSIS, FTTP so technologies marketed over 100Mbps.
BROADBAND GENERATION 2 = would cover FTTC so technologies marketed between 30Mbps to 100Mbps.
BROADBAND GENERATION 1 = would cover ADSL so technologies marketed up to 30Mbps.
While I accept this isn’t perfect it does present the information in an easy way for jo public, and me;) to understand.
Interesting but would still much rather list both as there are important points to be made
Dumbing down content is the opposite of what should be happening.
I get that there are important points with having both but it is Swahili to everyone except the most technical readers.
If your target audience isn’t jo public, MPs, Which readers, etc. then who, baring in mind that ISPReview write their articles in way for most of these people I’ve just listed to follow?
Don’t get me wrong Andrew you and your team have some of the best quality broadband data in the country that is highly technical and very detailed. So I’m hoping you will present your information in a way our leaders for instance in this case council leaders can easily digest and make the right decisions because many other third party reports of this nature, best and worst broadband areas, that appear on news websites like here and the BBC are mostly based on poor quality meaningless data.
@Alex Bristol
I do not think your categorisation makes any sense at all. Those “gen” categories are arbitrarily defined and don’t relate directly to speed.
Andrew’s stats are actually very clear and precisely defined. It’s much better they are defined by speed than technology coverage as the latter will be what’s meaningless to most people. All people need to know to understand Andrew’s tables are the difference between mean and median, the difference between upload and download speeds.
If it was to be put into more accessible terms, then those speed categories could be referenced in functional terms according to usage (such as how many 4K video streams).
One thing the mobile industry was very clever to do was market their technology as 1G to 5G which has greatly helped the public understand the technology and rather than saying EDGE we all say 2G because most people have a better idea of what that means, so it would seem a no brainer to apply the same wisdom to the broadband industry to help jo public better understand, and yes it is arbitrarily defined just like 1G to 5G is and yes I am sure you can find many other faults with it just like we could with the mobile 1G to 5G concept.
Steve your comment “Andrew’s stats are actually very clear and precisely defined.” and I wrote earlier “it is Swahili to everyone except the most technical readers.”, Steve you are a technical reader.
I’m a technical person myself and we are on a website that has many none technical readers and I feel there are parts of this Thinkbroadband report that are too technical and my view is also backed up by the evidence, where Andrew has had to clarify or correct several postings. Technical people like us should be making the technology easier to understand and less confusing to none technical people by simplifying things where possible which in turn helps improve our industry’s reputation with the general public.
Very misleading report indeed. In fact if you look at the full table Fermanagh and South Tyrone are bottom, West Tyrone third from bottom, Mid Ulster seventh from bottom and South Down ninth from bottom. Great to see Northern Ireland featuring so widely in the bottom 10. I remember when NI were leading the way with Superfast Fibre UK wide, looks like Openreach/BT Ireland have fled the country for some reason…..
Believe you are looking at the wrong table, hence you have a different order.
Report is based on local authority areas, the naming you are using matches Parliamentary Constituency naming and thus areas are of a different make up.
https://labs.thinkbroadband.com/local/fermanagh-and-south-tyrone,N06000007 36 Mbps
https://labs.thinkbroadband.com/local/west-tyrone,N06000018 40 Mbps
People saying Bournemouth is high because of Gigler is so far from the Truth. Gigler, mostly installs in ‘new development properties’ and big flats. They are just like BT, they will not install it for the masses, only what is convenient.
Who does install for the masses?