NAO reports on Superfast programme and warns of challenges for 2025 target
A new 52 page report from the National Audit Office simply titled Improving Broadband has seen the light of day today and looks at the Superfast broadband programme which saw £1.9 billion of public money (report estimates £0.9 billion will be returned under clawback clauses to local authorities). It also looks a little bit at the Gigabit 2025 programme and how this will work and the dangers that exist.
Despite wide coverage, many people in the UK still experience poor broadband. Suppliers to the Superfast Programme were encouraged to prioritise roll out to the easiest to reach premises, which meant premises in rural or remote areas were left behind. Rural coverage of superfast broadband is now at 80%, compared to 97% in urban areas, and is the lowest in rural Northern Ireland, at just 66%. Only 57% of UK premises that have access to superfast broadband are signed up to superfast packages as they may be unaware that faster services are available, may find their existing service sufficient, or consider faster services too expensive. Consumers may also not experience their advertised broadband speeds for several reasons including factors in the home.
The existing broadband infrastructure has been put to the test in the COVID-19 pandemic. Ofcom considers the existing infrastructure to have held up well, although some stakeholders representing areas with large rural populations told the NAO that the COVID-19 pandemic has exacerbated the rural-urban divide.
So far, the Department has prioritised increasing broadband coverage over speed. While superfast broadband is fast enough for most household use today, internet use is rising by 40% each year driven largely by video streaming and the Department is now focused on ensuring that the UK has a broadband network that will meet future demand. In 2018, to meet future demands of consumers and businesses, the government announced a new policy for the UK’s telecoms industry to provide infrastructure capable of faster gigabit speeds to 50% of premises by 2025, and nationwide by 2033. It later committed to accelerating this target and achieving nationwide coverage by 2025. It is still finalising its plans for its £5 billion Future Programme supporting delivery to the hardest to reach 20% of premises and has much work still to do and little time to do it. At present only 27% of UK premises are covered by gigabit-capable infrastructure.14% can access full-fibre, which is one of the lowest rates in Europe.Extract from NAO press release on Improving Broadband report
The 27% Gigabit coverage is actually our figure as Ofcom has not published any figures on Gigabit coverage, our figure merges together the full fibre coverage and Virgin Media DOCSIS 3.1 (Gig1 product). The 14% full fibre is sourced from Ofcom as are the majority of coverage figures in the report to maintain consistency and this figure is from May 2020, our figure for FTTP in May 2020 was between 13.5% and 14% and given Ofcom only reports at the integer resolution it seems we agree on the FTTP footprint. As of today 16th October the Gigabit coverage of the UK is 27.35% and FTTP at 17.28%.
The report makes it clear that the date deadlines for the first two phases of the superfast projects meant that to keep hitting local contractual targets and thus the over 90% and 95% targets the builds were prioritised to the easier to reach and VDSL2 was the preferred technology for its speed of roll-out.
The phase 3 and other little bits being using clawback money are almost all exclusively FTTP based and the report does acknowledge this change in what is being delivered. The downside to this is that the pace of the roll-out has slacked considerably, and there is now the expectation that some BDUK superfast work will continue into 2024.
The superfast programme is very much the past, and for all the complaints that BT ran off with the money once all the clawback has been resolved it looks like something like £1.2 billion of public money went into getting superfast (or better) speeds to 5.28 million premises. The negative side is that a good number of these 5.28 million will form the 10 to 20% of the UK that the Gigabit programme will need to subsidise to roll-out something Gigabit capable to.
On the Gigabit 2025 target it seems we might all end up calling it the Future Programme. An unfortunate title if it is the actual title since future tends to suggest something some time away and 2025 is racing up on us all very quickly.
The Future Programme is intending to award something like 1,400 different contracts as BDUK believes this will help to address some of the concerns that under the superfast programme the contract lots were too big and thus did not attract smaller operators. This also means if as happened with the Devon and Somerset phase two contracts that there are problems the impact of a few bad lots will be less.
The danger here though is that centralising a lot of the work will bypass teams working in the local authorities. On the surface for some local authorities when viewed from the outside it seems that once the contracts and photo opportunity was over that everything was left to BT whereas in others they are more pro-active and visible to the public. This visibility and interaction means in some cases there is a wealth of local knowledge, the Audit Office warns this knowledge may be lost particularly as councils adapt to the extremely fluid situation we are all living in. We fully agree with the Audit Office on this point but cannot see how councils will be able to allocate the staff time under the new model without some funding for local liaison.
Something we spend a lot of time on (and is starting late today as we've had a lot of reading to do for this news item) is mapping and some of what the Audit Office states is a surprise. The full address level data held by Ofcom is only from ten providers and Connected Nations has a further 14 providers but at the lower postcode level. This a surprise as we map at the postcode level 42 FTTP providers alone and there is Fibrus that has just been added to the database but is not on the public map yet.
3.5 In preparation for its Future Programme, where the Department will identify premises for intervention, it has started to produce preliminary lists of properties it considers eligible for intervention. The Department has provided this information to local bodies to validate. One local body told us that they had seen an error rate of between 40% and 70% in the data for their area. This was mainly because the data included properties which had already had full-fibre installed by an alt-net. Incomplete or incorrect data about existing infrastructure and supplier plans might lead to ineligible properties being selected for public subsidy.
3.6 Good data can help to maximise value for taxpayers’ money by enabling the Department to select an optimal balance of expensive and less expensive interventions. Its approach is to group individual premises into clusters, and then group clusters into contract bundles. It may choose to combine premises in a way that allows for cross-subsidisation within each contract. The way in which the Department determines bundles will be critical to successful delivery. The Department acknowledges that the data are imperfect and has told us that it recognises this is a potential risk. We have not seen the Department’s plans for mitigating this risk.
3.13 Under its new centralised model, the Department will need to replicate, at a national scale, the mapping and procurement functions performed by local bodies under the Superfast Programme. It expects many other functions supporting delivery and benefits realisation to remain within local bodies but, in practice, this will be subject to their ability to retain the necessary resources. It must identify and prioritise interventions for approximately 4.9 million premises and fund, procure and manage all 1,200 supplier contracts. The Department has identified that its existing systems, based largely around email and spreadsheets, will be inadequate to support this activity at the scale required. It intends to build a new system to receive, combine and process datasets from a variety of sources to build up the national picture and determine when and where contracts should be awarded. However, it will undertake this as part of a wider technology transformation programme. The Department acknowledges that it may not be ready in time for the anticipated start of the first contracts for the Future Programme in September 2021.NAO report on mapping risks
That DCMS/BDUK is using email and spreadsheets as the largest part of its current system is somewhat worrying, especially given that to map the UK at address level you are looking at some 30.5 million data points (i.e. UPRN) rather than 1.7 million postcodes. GIS systems are standard fare for providers planning and investors investigating whether to fund a roll-out, we built our own custom system around a SQL database starting back in 2012 and while we work at the postcode level (e.g. SW1 1AA) we can ingest URPN level data and have looked at working with UPRN. The problem we face in going full UPRN level is that with out adding several people to the team we would end up lagging the Ofcom reports whereas at present we are several months ahead of Ofcom. Full commercial access to UPRN data costs £20,000 a year, this is waived in some circumstances, but that we understand would impose restrictions on what we can do commercially with our results.
The warnings about missing the 2025 target i.e. 100% Gigabit coverage by 31st March 2026 are appropriate, we are generally more optimistic than most but that may be because we are able to see things improving daily. The issue will be that final 1 to 2% and these will also be those who missed out on the superfast programme, USC grants and USO intervention, i.e. those seeing £40,000 to £125,000 quotes for fibre on demand. The £4.5 billion (£500 million likely to be spent on vouchers) , just 10,000 properties costing £100,000 to serve with Gigbit would use up £1 billion of the fund and the £4.5 billion is meant to help between 3 million and 6 million properties.
The solution seems to be to water things down and deliver something that is not Gigabit but ultrafast for possibly 1%. The question then arises as to what happens in another five years once Gigabit becomes not just a nice to have but absolute need and 300 Mbps is feeling as painful to use as slow ADSL is today.
Ultimately what needs to happen is for the sticking plaster voucher schemes to be replaced by a comprehensive Gigabit roll-out and greater co-operation between DCMS, Ofcom, local authorities, commercial providers, community providers and people such as us and Point Topic who hold mapping data.
A couple of projections to end with based on our data and taken from our data portal after it finished its overnight analysis:
- If Virgin Media turns on DOCSIS 3.1 everywhere and no more FTTP is built the UK would be at around 60% Gigabit coverage, leaving 12.2 million needing FTTP intervention
- The projections for the UK reaching 15 million FTTP premises varies between Nov 2027 and Mar 2029 depending on whether build rate from last 3 months or 12 months is used.
- For UK to build a full 30 million premises projections range from 2036 to 2039.
So in theory building 12 million or so FTTP premises if built in the right places will get us to the target, but the build rates are very optimistic because the FTTP roll-outs are in volume terms very urban centric. The pace will slow down dramatically as you soon as you head outside the towns and villages i.e. venture into the final 5% or so of properties.
Politically the next year as Virgin Media turns on the DOCSIS 3.1 services means lots of celebration as the Gigabit coverage will increase quickly but once those easy pickings have happened the pace is going to look a lot worse.