This paper presents the outputs and briefly analyses the results of our take-up models at a granular level in the UK. We present the bulk of the analysis based on Lower Super Output Areas underpinned by availability data at a postcode level.
Actual take-up data, by operator and technology, is not available to us at a granular level. Our national level publications depend on announcements from the various players in the market, some who do not report directly useable data for the UK and as a result a number of data points are derived or modelled. See more in our UK Plus service.
To generate our estimates at LSOA level on down to postcodes we use a selection of modelling techniques to present a range of take-up and technology adoptions.
There has been a significant change in the broadband market in the last two years. The UK had been adding subscribers at a steady, declining, rate in the last half of the last decade and the transition from the FTTC based population to higher bandwidths was underway. Several factors have contributed to a slightly increased rate of take-up overall and a more noticeable shift to higher bandwidths and supporting technologies.
Virgin Media had led the way in terms of bandwidth offerings but as more and more alternative network players appeared (Gigaclear, Hypertopic, Community Fibre et al.) they were finding demand for full fibre and even gigabit services in their particular footprints.
The separation of Openreach and BT, and the subsequent strategic shift away from Gfast to an FTTP deployment focus as well as a regulatory shift to encouraging infrastructure rather than service competition and then the pandemic have all contributed to a rapid progression of what is on offer across the consumer landscape.
Increased demand for more bandwidth and more upstream capacity has meant that full fibre alternative networks in particular picked up subscribers quickly in 2019 and 2020.
Using speed tests across the thinkbroadband network we can provide another view of what the market shares are by operator and technology and some insight into what tariffs are subscribed to.
From a base of just under 14 million speed tests over the last 3 years we have selected the fixed retail tests (as opposed to satellite, leased line and mobile) from 2020. This results in a dataset of 7.1 million speed tests by postcode across the UK with up and down bandwidths, technology and operator.
In the case of FTTC we found that in 2020 the proportion of FTTC subscriptions that tested under 30Mbps was 22.4%. Those users are therefore on ‘standard’ broadband.
It is important to note the biases inherent in this data. Primarily the universe of tests is self selecting and tends to be from users who are happy or unhappy with their services.
It does not help us with which overall take-up or totals and the actual numbers of subscribers.
The data does offer support for modelling. We can draw some confidence by comparing our own data from operator announcements with the speed test analysis. Note that the real world data is at a national level from December 2020, see Superfast and ultrafast lines reach 80 per cent share milestone–an analysis of Q4 2020 UK ISP numbers | Point Topic subscribers (point-topic.com) and the speed test data is from all of 2020.
If we review the market shares from speed tests against our national data, from our UK Plus service, we can see broad alignment.
One interpretation is that those on ADSL and FTTC are testing more since they are unhappy with their bandwidths while those on DOCSIS are relatively comfortable.
The more granular the analysis of the speed tests the more likely there will be local divergence from ‘real world’ data. However without access to granular outputs on subscribers and technologies we believe this analysis has value and valid applications for local modelling.
We can do the same for ISP market shares again checking against our national data.
A more complicated picture giventhe sheer number of ISPs in the analysis, which will skew the data particularly given the lack of reporting of subscriber numbers from the smaller players.
You can then cross tab the data further to look at ISP and technology market shares. Note again that as the data is broken down by dimensions and geography the smaller the sample becomes and the more likely there is to be variance from the real world situation.
The latest collaboration between Point Topic and thinkbroadband has resulted in a new product set, ThinkPoint. Two of the most respected industry analysts and data specialists offer the most complete, up to date, granular and accurate representation of where broadband is today, from who, provided how and what the current and future picture of supply and adoption is.
We are now able to offer more data, with more granularity and the most up to date view of broadband in the UK from postcode to national level. Our data and analysis are available from single fields at a local level through to the complete national picture.
For more information please email firstname.lastname@example.org