uk broadband mapping -...
Post on 28-Jun-2019
219 Views
Preview:
TRANSCRIPT
UK Broadband Mapping:
Methodology for availability and take-up mapping
Prepared by: Oliver Johnson
Date: 2013
Point Topic Ltd 73 Farringdon Road
London EC1M 3JQ, UK Tel. +44 (0) 20 3301 3303
Email oliver.johnson@point-topic.com
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 2 of 27
Contents
1. Introduction _________________________________________________ 3
2. The foundations of broadband mapping __________________________ 4
3. Mapping exchange properties __________________________________ 5
4. Mapping the speed of ADSL broadband _________________________ 10
5. Mapping the availability of superfast services ____________________ 12
6. Mapping the availability of cable services _______________________ 14
7. Mapping the availability of FTTx services ________________________ 16
8. The roll-out of next generation access services____________________ 19
9. Broadband take-up by postcode _______________________________ 22
10. Operator market share across the UK ___________________________ 24
11. Digital deprivation in the UK __________________________________ 26
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 3 of 27
1. Introduction
Every six months, Point Topic updates its database of broadband availability and take-up in every
postcode in the UK. We report on key broadband coverage and take-up metrics for the UK’s fixed
broadband market:
Exchanges – we map the area served by each telephone exchange in the UK. We report
whether the exchange has been unbundled, which operators offer services and the
technology installed at the exchange.
Cable coverage area – we map the full cable franchise area and our estimate of the
postcodes that are currently activated;
FTTx availability – we look at the exchanges that have been enabled for fibre and estimate
the likely coverage of households in those exchanges. We also track coverage of the UK’s
alternative fibre providers;
Downstream speeds – typical modem sync speed and throughput speed for copper services
and top-speed available from any broadband service;
Superfast forecasts – we report how likely is it that premises within the postcode will be
able to access superfast services in the next five years.
Standard and superfast take-up – we report for every postcode how many standard and
how many superfast lines are adopted
Operator market share – the models also allow us to split the standard and superfast take-
up numbers by major operator. This leads to market share statistics for each post code in
the UK and all the geographies above.
Residential and business take-up – we track and update the number of business premises
around the UK using inputs like the Royal Mail code point updates and the Valuation Office
data on hereditaments in the UK
This document describes how we create our databases for UK broadband availability and take-up.
If you have any questions, or if you would like to access our full database of broadband availability
and take-up in every UK postcode, please e-mail me at oliver.johnson@point-topic.com. .
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 4 of 27
2. The foundations of broadband mapping
2.1 A unit postcode database
Our datasets are built from a database of every unit postcode within the United Kingdom. Point
Topic uses the Code Point postcode database published by the Ordnance Survey to provide the
precise geographical location of the centroid of each postcode.
2.2 Households, business numbers and population
Household, business and population numbers are published regularly by national government
organisations1, usually at the local authority level. This data needs to be disaggregated down to the
postcode level.
Ordnance Survey publishes the number of domestic and non-domestic drop points within every
postcode in the UK. We use this to estimate the distribution of households, businesses and
populations within every postcode in the UK, based on the aggregated data published by the
national statistics offices.
2.3 Higher geographies
Each unit postcode is allocated to higher level geographies:
Country, Government Office Region, Local Authority, MSOA, LSOA, telephone exchange, postal
sector, Census Output Area (COA).
Our UK Broadband Mapping data is available for every postcode in the UK, although some users may
choose only to purchase the higher level or aggregated datasets. These datasets are most useful for
making quick comparisons between areas.
1 Households and population: DCLG on gov.uk, ONS, Stats Wales, General Register Office for Scotland.
Business premises: Valuation Office Agency, Scottish Assessors Association, Department of Finance and Personnel (Northern Ireland)
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 5 of 27
3. Mapping exchange properties
To map the availability of broadband across the UK, it is essential to have an understanding of the
locations and boundaries of its telephone exchanges. BT publishes the postcode location of its
exchanges, but is yet to provide a publically available source of data for the area served by each
exchange.
Point Topic has therefore developed a methodology to map the boundaries of each exchange and
therefore map each postcode to its local telephone exchange.
3.1 Mapping the exchange boundaries
The postcode location of all exchanges is published, but to understand broadband availability we
need to be able to map the boundaries of the exchanges.
Point Topic has therefore modelled the likely boundary of exchanges. Our estimates have been
shown to be 87% accurate, with the greatest inaccuracy lying on the boundaries of exchanges and in
postcodes that are served by multiple exchanges.
Our approach assumes that:
• An imaginary line connecting any two neighbouring exchanges will be bisected at a right
angle by an exchange boundary.
• The exchange boundary will be exactly midway between the two exchanges.
Figure 1: Voronoi tessellation is
used to estimate the KCOM and BT
exchange boundaries
We use the same approach to estimate the boundaries of KCOM Group exchanges within the UK. In
this way, every postcode in the UK is allocated to its local exchange.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 6 of 27
3.2 LLU operators present within an exchange
Point Topic uses data published by SamKnows to identify which exchanges have been unbundled for
LLU operators. Combined with our data on the exchange boundaries, we can then calculate the
coverage of LLU operators.
Local loop unbundling in the UK is still on the rise, although its growth rate is slowing. By the end of
June 2013, 92.4% of homes and businesses in the country were within reach of LLU broadband
services, up from 89% at the end of 2011 and 84% at the end of 2010.
The proportions vary by country – with 85% of premises covered in Northern Ireland and Scotland,
87% in Wales and 93% of premises covered in England.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 7 of 27
Figure 2: LLU operator presence end June 2013
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 8 of 27
3.3 Technologies present within an exchange
The technology installed in the exchange can affect the service its users experience.
In every exchange, we report whether ADSL Max or 21CN (“twenty-first century”) technology has
been installed.
ADSL Max
In those exchanges that don’t have ADSL Max, broadband speeds would never be expected to
exceed 1.5Mbps.
21CN network
BT Wholesale has met its target to WBC-enable exchanges serving around 90% by Spring 2013,
thereby bringing the advantage of 21CN for areas with limited broadband infrastructure including
higher speeds. We expect retail broadband prices in areas 21CN-enabled areas to see a reduction,
although those without LLU are likely to remain higher.
3.4 Distance from the exchange
One of the most frequently asked questions about the broadband environment in the UK is about
the profile of local loop lengths.
The length of the pair of wire (or the loop length) which connects each telephone end-user to their
telephone exchange is one of the main factors which determine the data bandwidth the customer
can be provided with over a DSL service. So the profile of loop lengths reveals, for example, how
many homes can be provided with a fast enough service to stream video content using a DSL
connection?
One reason the question is asked so often is that the information is not readily available. In 2005 the
Chief Technical Officer of BT Exact, the telco’s research and development unit, revealed the broad
profile of loop lengths in the UK. From this presentation, Point Topic has extracted the relationship
between the linear distance from the exchange and the actual loop length.
The average loop length in the UK is around 3.4km, and the majority (27%) have a loop length
between 2 and 3m. Distance from the exchange is a key factor that helps us determine the speed of
an ADSL broadband connection, as described in the next section.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 9 of 27
Figure 3: the loop length profile of the UK - % premises by local loop length. Source – Point
Topic, June 2013
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 10 of 27
4. Mapping the speed of ADSL broadband
There are many factors which can affect the speed a user receives when connecting to the internet
in their home, not least the tariff that they subscribe to.
Assuming that all users subscribe to the highest available tariff, the single most important factor
which affects the data bandwidth a user can receive over an ADSL service is the length of the pair of
wires (or the loop length) which connects each telephone end-user to their telephone exchange.
Point Topic has completed a statistical analysis of 1.6m speed test results, to establish the
relationship between ADSL speed and the implied local loop length.
Using speed test results alone is not a robust way to estimate the likely speed in an area. There are
many different factors that can affect speed at any particular time – contention, tariffs, backhaul
provision – and these will all be reflected in the reported results. Getting close to the average speed
that users in a postcode should expect to get is therefore very difficult and requires large sample
sizes for each of the UK’s 1.7m postcodes. Point Topic’s approach – modelling based on speed test
inputs – produces the most accurate estimates per postcode available publically.
We report two key measures of speed for ADSL broadband services, based on our analysis of speed
test inputs:
Modem sync speed (@75%)
Modem sync speed is set by the ISP. This is an optimal number since:
a) this is only the speed between you and your ISP with nothing in between slowing the
signal down;
b) when you're actually sending information to and fro over the internet you have a certain
amount of bandwidth that has to be allocated to packet routing.
For every postcode in the UK, we report the modem sync speed that users should expect to
get 50% of the time. The actual speed that they receive could then be higher or lower.
Throughput speed (@75%)
The throughput speed is a closer reflection of the speed that a user will actually experience.
We provide the average bandwidth which users should expect to get 75% of the time,
assuming that they subscribe to an up to 24Mbps (maximum) ADSL2+ bandwidth tariff from
an ISP.
The technology available at an exchange is also an important factor that can limit ADSL speed.
Premises in exchanges without 21CN technology should not expect speeds above 12Mbps. Premises
in exchanges without ADSL Max should not expect speeds above 1.5Mbps.
The chart below shows the range of throughput speeds that premises in the UK can expect to
receive 75% of the time. At the end of June 2013:
0.6% of premises (180k) could not typically access speeds above 0.125Mbps – technically
not broadband;
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 11 of 27
12.1% of premises (3.5m) could not typically access speeds above 2Mbps. In terms of actual
application, this means that they would not be able to get a good service from online
streaming services;
9.2% of premises (2.7m) can typically access speeds above 12Mbps. These are the premises
which are closest to their local telephone exchange.
Figure 4: the ADSL throughput speed profile of the UK. The chart shows the range of ADSL
speeds that users should expect to receive 75% of the time. Source – Point Topic
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 12 of 27
5. Mapping the availability of superfast services
Up to this point, we have only discussed our methodology for establishing the coverage of ADSL
broadband providers. The rest of the document now focuses on broadband delivered over cable and
fibre, or next-generation access (NGA) technologies – so called because they can potentially deliver
much faster speeds to the consumer.
The geographical density of demand is by far the biggest single factor shaping the economics of
NGA. In simple terms:
The cost of providing NGA is roughly proportional to the size of the area covered;
The return is proportional to the number of customers served.
Areas with more people per square kilometre are therefore more likely to get NGA on a commercial
basis. Less dense areas will depend on some degree of subsidy or special assistance to get NGA.
A further factor is the variation in demand for services in different areas. There can be significant
variations in demand depending on the socio-economic mix within an area and the number and type
of businesses.
Point Topic uses the ‘demand density’ in an area to understand the roll-out of superfast services
within the UK.
Demand density = Expected broadband demand / size of area
We divide the whole of the UK into nine demand density bands. In the lowest band the potential
broadband demand is under three lines per square kilometre. In the highest band the potential
broadband demand is over 3,000 lines per square kilometre.
In November 2012, Ofcom published a database of superfast availability per postcode, reflecting the
market in June 2012. Using this data, we can then profile the coverage of superfast services by
density band. As expected, we see a clear skew towards the highest density bands, where nearly
75% of premises are able to access superfast services. In the lowest band, around 6% of premises are
able to access superfast services.
Clearly then, the density of demand is an important factor which determines superfast deployment,
and Point Topic uses this to estimate likely coverage in every UK postcode and to forecast future
patterns of deployment.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 13 of 27
Figure 5: superfast coverage in each of Point Topic’s nine demand density bands. The results
show a clear skew towards areas with high demand density
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 14 of 27
6. Mapping the availability of cable services
To estimate cable coverage in the UK, we use a combination of publicly available inputs, proprietary
data and modelling.
Figure 6: Point Topic’s approach to mapping the coverage of Virgin Media’s cable services
Virgin Media’s franchise area is published, and they announced the total number of homes passed
by their network – 12.5m homes at the end of June 2013. We estimate that in total Virgin Media
passes 13.4m premises when including businesses. Point Topic then uses speed test data inputs
from Ofcom and ThinkBroadband to establish where within the franchise area Virgin Media is
providing services.
Using this approach, we are able to identify 6.5m premises passed by Virgin Media. It also allows us
to look at the typical deployment patterns of the operator, in terms of the types of premises that its
network targets.
Virgin Media specifically targets the higher (but not highest) density areas within its franchise area.
Using this insight, we are then able to estimate the likely distribution of the rest of its coverage.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 15 of 27
Figure 7: Virgin Media coverage in each of Point Topic’s nine demand density bands. The
results show an even more pronounced skew towards areas with high demand
density, with the lowest density areas receiving very little coverage
This model has been shown to be over 90% accurate when compared to actual deployment by Virgin
Media. The greatest inaccuracies occur when a postcode is only partially served by services.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 16 of 27
7. Mapping the availability of FTTx services
7.1 BT’s Infinity Service
BT Openreach regularly publishes which exchanges will be enabled for its fibre-based services. This
data is used by Point Topic to identify exchanges which have been enabled for fibre, will be enabled
for fibre and which technology is installed.
Having an upgraded exchange doesn’t necessarily mean that homes and businesses within the
exchange area will be able to receive superfast services. At the end of June 2013, BT announced that
its fibre-based network passed over 16 million premises – out of a total of 20.7 million premises in
enabled exchanges. This means that on average around 75% of premises are activated within the
enabled exchange.
To receive fibre broadband services, BT must upgrade both the local exchange and street side
cabinets. If your cabinet has not been upgraded, you won’t be able to access the services.
Since BT does not release full datasets on the location of its cabinets or the cabinets which are
enabled, it is very difficult to get an independent view on the coverage of BT’s superfast services
across the UK. BT has stated that the decision to upgrade a cabinet is based on the infrastructure
and the market within the area served by the cabinet.
Again, Point Topic’s approach to estimating the precise coverage of BT’s fibre-based network relies
on a combination of publicly available inputs, proprietary data and modelling.
BT Openreach publishes which exchanges will be enabled and when. Using Point Topic’s exchange
boundaries, we can then map the maximum potential coverage area of fibre based services.
We then use speed test data inputs from Ofcom and ThinkBroadband to establish where within the
exchange areas BT Openreach is providing services. This allows us to establish more than 50% of BT
Openreach’s postcode coverage with a high level of certainty.
Again this allocation can be used to establish BT Openreach’s typical deployment, in terms of the
density of areas targeted. We then use these typical deployment patterns to estimate where BT’s
remaining coverage is likely to be across the activated exchanges. Infill forms an important part of
BT’s fibre roll-out, and so the age of an activated exchange is also an important factor to consider.
We see very different patterns of deployment, depending on whether the network is competing
with cable services and whether it is a commercial or state-backed project.
7.2 Commercial roll-out outside the cable coverage area (UK profile)
When BT Openreach is not competing directly against Virgin Media, it appears to target areas of the
UK which are in the middle of the demand density profile.
Note that this profile excludes the Cornwall and Northern Ireland projects, which were state-funded
and exhibit a different demand profile.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 17 of 27
Figure 8: BT’s FTTx coverage in each of Point Topic’s nine demand density bands – without
competition from Virgin Media. Source – Point Topic
7.3 Commercial roll-out within the cable coverage area (UK profile)
However, when BT Openreach’s network is competing directly with Virgin Media, it shows a much
stronger skew towards the areas with the highest density of demand and therefore revenue
potential.
Figure 9: BT’s FTTx coverage in each of Point Topic’s nine demand density bands – with
competition from Virgin Media. Source – Point Topic
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 18 of 27
7.4 State-funded roll-out
Cornwall and Northern Ireland continue to showcase BT’s approach to state-funded roll-out.
The deployment pattern is again very different still and highlights the requirement to serve areas
which would not receive funding on a commercial basis.
Figure 10: BT’s FTTx coverage in each of Point Topic’s nine demand density bands –with state
funding. Based on deployments in Northern Ireland and Cornwall. Source – Point
Topic
7.5 Using the density demand profiles
The Ofcom and ThinkBroadband data inputs allowed us to identify about 50% of BT’s fibre-based
coverage.
We use these density demand profiles to estimate how likely it is that other properties in the
exchange are activated, based on:
The expected the demand density in the area;
Whether BT is competing against Virgin Media;
Whether the postcode is in an exchange upgraded through BT’s commercial roll-out or a
state-funded programme;
How long the exchange has been fibre-enabled.
The last point is important. Infill forms an important part of BT’s fibre roll-out, and cabinets in an
exchange will continue to be upgraded after the exchange is upgraded.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 19 of 27
7.6 Alterative network fibre providers
Point Topic provides the most detailed assessment of the coverage of alternative network or ‘altnet’
fibre providers available.
At the end of June 2013 we tracked 28 active altnet fibre suppliers across the UK. We estimate that
these providers covered around 3% of premises in the UK. Whilst coverage is currently limited, these
providers are expanding their coverage and offering the fastest speeds available in the UK at the
moment.
7.7 Top-speed available
The top speed available to an area depends on the technology that is available.
Figure 11: Top download speed available in the UK, % of premises. Source – Point Topic
At the end of June 2013, 0.1% of premises were able to access gigabit services from Gigaclear,
Hyperoptic, B4RN or CityFibre. 5.3% were able to access BT’s 330Mbps FTTP services or KC’s
350Mbps service. For 44% of premises cable is their fastest service available, and for 27% their
fastest service is BT’s 80Mbps FTTC service. The rest of the UK’s premises can only access copper
based broadband services.
8. The roll-out of next generation access services
Point Topic forecasts the likely roll-out of superfast services in the next five years. Our forecasts use
targets set by government, BT and the BDUK projects to estimate which postcodes are likely to get
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 20 of 27
coverage and when. It allows us to identify the areas that will still be left behind after the
commercial and state-funded roll-outs are complete, or the ‘final 10%’.
It is important to note that our new forecasts assume that the targets for roll-out will be met. This is
an area of considerable debate. The model does not attempt to foresee whether targets will be met
– instead it focuses on the areas that will be left behind once different levels of coverage are met.
8.1 UK superfast broadband targets and milestones
At the end of 2012, superfast services were available to 70% of households in the UK. This was one
of the published findings from Point Topic’s study for the European Commission – Broadband
Coverage in Europe in 2012.
BT has announced that it is on target to complete the commercial roll-out of its superfast services by
spring 2014. There is no indication at present that Virgin Media plans to extend its superfast
network.
In June 2013 the UK government announced a revised target to bring superfast broadband to 95% of
the country by 2017. This effectively replaced the previous aim of 90% superfast coverage by the
end of 2015.
Most of the BDUK programmes have now announced their coverage targets and the start and end
date of the roll-out programmes.
Point Topic has used these announcements to forecast the likely roll-out of superfast services within
the UK.
Whether targets will be met is an area of considerable debate. On 5 July 2013 the National Audit
Office (NAO) published a report saying that the Government programme to make superfast
broadband available to 90% of premises in the UK is expected to be delivered two years later than
initially planned. They also predicted that only nine out of 44 local projects were expected to reach
their original target of providing 90 per cent superfast coverage by May 2015.
The focus of this analysis is not to question whether the targets will be met on time. Rather it is to
focus on who will be left behind once targets are met, or the ‘final 10%’.
8.2 BT’s commercial roll-out of superfast broadband
In February 2013, BT announced the final 99 exchanges that will be upgraded to receive superfast
broadband services as part of the commercial roll-out of its services. Then in July 2013 the network
operator announced that nine new exchanges will be enabled for FTTC services, essentially
squeezing these in under its commercial rollout. In total, BT’s commercial roll-out plan is expected to
bring 80Mbps Fibre-to-the-Cabinet (FTTC) or 330Mbps Fibre-to-the-Premises (FTTP) services to over
2,000 exchanges by the spring of 2014.
At the end of June 2013, BT’s commercial roll-out had brought superfast broadband to over 16
million premises (homes and businesses). Its target for 2014 is to cover 19 million premises, or 66%
of all UK premises.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 21 of 27
All exchanges within the commercial roll-out have now been announced through the BT Openreach
website, but no data is publically available on precisely which cabinets will be activated.
As we demonstrated in Section 8, BT’s deployment varies depending on whether it is facing
competition from Virgin Media. We use these deployment patterns to estimate the likely roll-out of
BT’s commercial deployment, assuming that they meet their target to deliver fibre-based services to
66% of premises by spring 2014.
8.3 BDUK state-funded roll-out of superfast broadband services
At the end of July 2013, 29 local authorities had signed contracts with BT. A further 15 local
authorities had yet to sign contracts.
The roll-out of state funded services will target different areas to commercial programmes. We use
the deployment profile seen in Northern Ireland and Cornwall to establish the likely roll-out of the
BDUK projects.
Our forecasts for the roll-out of state-funded services assume that:
The roll-out will follow a similar pattern to deployment in Northern Ireland and Cornwall;
The roll-out will not occur in areas already covered by Virgin Media;
BDUK projects will start and end on time and reach their coverage targets.
8.4 Predicting the roll-out of superfast services in the UK
Point Topic has created a model which looks at how likely it is that every postcode in the UK will get
superfast services. It looks at the likely roll-out of BT’s commercial services, as well as expected
BDUK investment.
Our new forecasts assume that the targets for roll-out will be met. This is an area of considerable
debate. The model does not attempt to foresee whether targets will be met – instead it focuses on
the areas that will be left behind once different levels of coverage are met.
The maps on the next pages show our forecast of superfast coverage at the end of 2012 and 2015.
Overall the UK is expected to have just over 90% coverage at the end of 2015 - Wales and the South
West will have the lowest coverage, with 83% and 84% superfast coverage respectively.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 22 of 27
9. Broadband take-up by postcode
We now have the following inputs for our model of broadband take-up in every UK postcode:
From the Ofcom published datasets, we know the take-up of standard and superfast
broadband services across the counties in the UK.
From Point Topic’s own research programmes, we know the total consumer and business
subscriber numbers for all major and smaller operators within the UK and their coverage
areas.
From our surveys, we understand the likely take-up of broadband by different types of
households and businesses within the UK.
From our research, we know which operators are active in which exchange and which
exchanges have been enabled for FTTx services.
We also have an estimate of Virgin Media’s coverage layer.
An overview of how we calculate broadband availability by postcode is provided in Appendix One.
We use these inputs to estimate the number of consumer and business subscribers for each
operator in every UK exchange. The process for estimating operator market share in each exchange
is as follows:
1. Calculate the number of premises passed by each operator within an exchange. Virgin
Media will not of course pass all premises within an exchange.
2. Estimate the expected number of broadband lines within the coverage area of each
operator within the county. This is based on the likelihood that the premises passed will
take-up broadband services.
3. Model One – distribute total operator lines for the UK by exchange according to the
likely take-up within the coverage area of each operator.
4. Model Two – adjust the distribution of operator lines so that the total number of
subscribers for each operator is as close as possible to the estimated county figures for
UK take-up.
We use this methodology to estimate the numbers of consumer and business, standard and
superfast lines for every operator in every UK exchange.
Operator subscriber lines are then distributed across every postcode within an exchange, based on
the expected take-up of broadband within the coverage area for that operator.
9.1 Broadband take-up in the UK Point Topic provides its subscribers with full databases of broadband availability and take-up across
the UK.
In the year between June 2012 and June 2013 for example, nearly 2.5 million superfast lines were
added in the UK market. Nearly 60% of these net additions were allocated to Virgin Media, with the
remainder allocated to BT and its resellers.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 23 of 27
Figure 12: Changes in superfast lines by operator. Source – Point Topic
Whilst BT has been expanding its network rapidly over the course of the year and now passes more
premises with superfast services than Virgin Media, Virgin Media’s superfast customer base has
grown as rapidly as BT Infinity’s within the year. Virgin Media is achieving this in part by upgrading
its existing customer base from lower speed to superfast services. There is also significant overlap
between the two networks at present, concentrated within the high population density urban areas
favoured by Virgin Media.
As a result, take-up of superfast services is still concentrated within Virgin Media’s coverage area.
The key exceptions are Northern Ireland and Cornwall, which have already benefited from state
investment in superfast broadband into semi-urban and rural areas. Adoption rates here are
encouraging for Broadband Delivery UK (BDUK) and operators (mostly BT) which will be expanding
into these areas in the future.
However the current focus on delivery of superfast services to all does not address those who still do
not take any broadband service, or non-adopters. The barriers to broadband adoption – not just
physical availability but also the desire and financial ability of people to take-up services – will
remain the biggest challenge for the UK government in the future. Point Topic’s work on the factors
that affect ‘digital deprivation’ is addressed at the end of this report.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 24 of 27
10. Operator market share across the UK
Our maps allow us to evaluate the expected market share of operators within any UK postcode. In
this section, we look at national and regional operator market shares.
Figure 13: Retail market share for operators by country. Source – Point Topic
Figure 14: Retail market share for operators by region. Source – Point Topic
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 25 of 27
The table below shows the shifts in operator market share between June 2012 and June 2013 by
region. Sky has experienced the biggest growth in subscribers over this period, growing by 23% in
the year, boosted by the acquisition of Telefónica UK’s O2 and its Be consumer broadband and fixed-
line telephony business completed on 30 April 2013. As the table below shows, Sky has increased its
market share in every UK region. In the East Midlands, North West and Wales, this seems to have
been mostly at the expense of BT.
In the same period BT’s subscriber base grew by 7%, Virgin Media by 2% and TalkTalk by 1%.
TalkTalk’s market share only grew in the East Midlands and Wales. Virgin Media experienced its
biggest gains in the East Midlands, Scotland and Yorkshire and the Humber.
Area Sky BT Retail Virgin Media
TalkTalk Other
East Midlands 6% -8% 3% 1% -1%
North West 5% -4% 1% 0% -2%
Wales 4% -6% 1% 1% 1%
West Midlands 4% 2% -2% -1% -3%
East of England 3% 2% -1% -2% -2%
South East 3% 3% -3% 0% -4%
London 3% 2% -1% -2% -2%
South West 3% 1% 0% -2% -2%
Scotland 2% 0% 2% -1% -3%
North East 2% 2% 0% -1% -4%
Northern Ireland 2% 2% 0% -2% -2%
Yorkshire and The Humber 1% 0% 4% -2% -3%
Figure 15: Changes in retail market share for operators by region, June 2012 to June 2013. Source –
Point Topic
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 26 of 27
11. Digital deprivation in the UK
Now the UK is moving towards 95% per cent fixed ‘superfast’ coverage before the end of the decade
attention has been shifting to those for whom the physical availability of a network is not the only
barrier to adoption.
Broadband digital deprivation is a looming challenge for the UK. The focus on broadband coverage
has, quite properly, dominated the conversation for the last few years. Now though the actual take-
up of a service is of as much concern when it comes to getting the UK online.
New research published by Point Topic has identified seven primary predictors of low adoption:
Age – the older an individual the less likely they are to have a broadband connection at
home;
Income – higher income means higher adoption rates;
Disability – the ONS has provided data indicating that half the non-adopting population are
registered disabled;
Housing – those in social housing also show low broadband adoption rates;
Availability – having better and more numerous broadband options drives higher broadband
penetration. Point Topic’s Broadband Infrastructure Index, which measures the quality and
availability of broadband services in an area, provides the input for this component;
Children – the presence of children in a household drives up broadband adoption;
Education – the higher the educational attainment the higher the rate of broadband
adoption.
Point Topic has developed a Broadband Digital Deprivation Index (BDDI) to highlight those areas
most at risk from internet/broadband non-adoption. The map below shows the risk associated with
each local authority in the UK.
The Broadband Digital Deprivation Index gives us a view of the factors affecting broadband adoption
and where their impact is expected to be felt. This is an important step in being able to address
those particular aspects.
The full report is available now to Point Topic’s UK Plus subscribers. A dataset of the Index in every
LSOA area for England is available to UK Broadband Mapping subscribers.
UK Broadband Mapping: Methodology for availability and take-up mapping
Page 27 of 27
Figure 16: Broadband Digital Deprivation Index for each local authority in the UK. Source – Point
Topic
top related