UK's £5bn Project Gigabit Hits 40% Completion - What It Means for Businesses Still Waiting for Fibre

The latest monthly progress figures from Building Digital UK (BDUK) show that the government’s £5bn Project Gigabit scheme has now delivered a gigabit–capable connection to 340,880 of the 861,890 premises it has under contract — a completion rate of 40%, up from 318,000 premises, or 37%, in September 2026. On the national numbers that is steady progress. Under the national numbers it is anything but. One Openreach contract covering North Herefordshire, North Wales, Shropshire and South West Wales is just 7% complete and added no premises at all this month. A Wildanet contract in Central Cornwall has finished. One Openreach contract in Devon and North Somerset went backwards.

For a UK business, the national percentage is close to meaningless. What matters is the contract that covers your postcode, its supplier, and how it is actually performing — and on today’s figures, 24 October 2026, the gap between the best and worst contracts is the gap between a business that can plan around fibre arriving next quarter and one that may still be waiting years from now. For organisations in the slower intervention areas, which BDUK’s figures place largely in Wales, Scotland, Shropshire and parts of South West England, the sensible reading of this month’s update is not “fibre is coming”. It is “design the network you can rely on today, and treat fibre as an upgrade when it lands”.

What BDUK’s October update actually shows

Project Gigabit exists to reach the premises that the commercial market will not reach on its own — the “final 10–20%” of the UK, typically rural, remote or otherwise expensive to build to. Commercial operators have already taken gigabit–capable networks to around 91% of premises, according to Ofcom, and the regulator forecasts that commercial build alone, independent of this scheme, will reach 95% by January 2029. The subsidy programme is aimed at what lies beyond that commercial horizon, with a goal of roughly 99% gigabit coverage nationwide by 2032. BDUK lets that work out as a set of regional contracts, each awarded to a single supplier, and publishes each contract’s progress monthly.

This month’s national total of 340,880 connected premises against 861,890 contracted represents 22,880 premises added in a month. If that pace held exactly — and the contract–level data suggests strongly that it will not hold evenly — the remaining 521,010 premises would take the best part of two more years to reach. That is simple arithmetic rather than a forecast, and it is useful mainly as a sense of scale: even on the national average, the programme is closer to the start of its work than the end.

The more instructive part of the update is the contract–by–contract spread. Wildanet’s Central Cornwall contract has reached 100%. CityFibre’s contracts in Cambridgeshire and in Bedfordshire, Northamptonshire and Milton Keynes are each at 82%. At the other end, Openreach’s CO3 contract — North Herefordshire, North Wales, Shropshire and South West Wales — stands at 7% and added nothing this month, while Openreach’s CO4 contract covering South Devon, Mid Devon and North Somerset recorded a net change of −10 premises, attributed to a pending data revision. Contracts in parts of Wales, Scotland, Shropshire and South West England remain under 20% complete. The same programme, the same month, and a range of outcomes wide enough that the national figure describes almost nobody’s actual position.

How Project Gigabit got here, and where it is heading

The programme’s headline dates stretch to the next decade. The dates that matter for an individual business are the contract milestones in its own region, and those are published monthly.

The same programme, very different results

The bars below set the national completion figures alongside the individual contracts named in BDUK’s October update. The spread is the story: a business in Central Cornwall and a business in Shropshire are covered by the same scheme and are in completely different positions.

Two things follow from that chart. The first is that delivery speed in this programme is a property of the individual contract, not of the scheme. The contracts at 82% and 100% demonstrate that the build can be done at pace in subsidised areas; the contracts below 20% demonstrate that it is not being done at pace everywhere. A business cannot infer its own timeline from the national figure, from a neighbouring county, or from what it read about the programme last year. It has to look at its own contract.

The second is that the lagging contracts are geographically concentrated in places where alternatives are also thinner. Rural Wales, Shropshire, parts of Scotland and the South West are not just where subsidised fibre is slowest; they are also where mobile coverage can be patchy, where a second fixed–line provider is less likely to be available, and where an engineer visit after a fault takes longer to arrange. A business there carries more connectivity risk on today’s infrastructure and has fewer easy routes to reduce it, which is exactly why it cannot afford to wait for a fibre date before thinking about resilience.

Most of the work is still ahead

Five years into a programme with a 2032 end date, the national figure is a fair indication of where things stand overall — and of how much of the build is still to come in the places that are hardest to reach.

It is tempting to read 40% as “well under way”. A more useful reading is that the programme is selected towards difficulty: Project Gigabit only covers premises the commercial market declined to build to, and within that group, the premises still waiting are the ones that were not easy enough to connect first. In practical terms that means long rural spans, difficult terrain, wayleaves across private land, poles that need replacing and duct that does not exist. The final stretch of any build of this kind tends to be slower per premises than the opening stretch, not faster, and month–to–month totals that occasionally go backwards as records are revised are what that complexity looks like in the data.

For the businesses inside that remaining 60%, the shape of the next few years is therefore uncertain in a specific way. Fibre will very probably arrive. The date on which it arrives, and how firm any published date turns out to be, is not something the business can control or rely on. That is a planning problem rather than a technology problem, and it has a well–understood answer: build resilience into the connection you have now, choose equipment that will carry over when fibre lands, and avoid long contracts that lock you into today’s product past the point where something better becomes available.

There is a timing collision to keep in mind as well. The national copper switch–off completes on 31 January 2027, and businesses in areas without fibre are generally moving from analogue lines to SOGEA–based broadband with hosted voice over the same copper infrastructure. That is a perfectly sound interim step, but it concentrates more on a single circuit than before: voice, card payments, alarm signalling and every cloud application now depend on one broadband line that, in a lagging intervention area, may be a long copper run with a modest speed and no alternative path. The switch–off makes resilience more important in exactly the places where the fibre build is slowest.

Where waiting on a single fibre date leaves a business exposed

The list below reflects the connectivity weak points most commonly found in UK SMEs located in slower Project Gigabit areas, with badges indicating how much attention each typically needs.

The pattern across the top three rows is concentration. Each one is fine in isolation; together they describe a business whose voice, revenue, safety signalling and data protection all fail at the same moment, for the same reason, with no way round it. That is the situation in which a fibre date becomes a liability rather than a plan, because it gives everyone a reason to defer the work that would have broken the dependency. The backup point deserves particular attention: on a slow upload link, a cloud backup that never quite finishes is indistinguishable, from the dashboard, from one that does — until the day it is needed.

What resilience costs while you wait

The bands below are indicative planning figures for UK businesses adding resilience to an existing connection pending fibre, covering hardware, installation and typical monthly service costs. They are not quotes. The variable that moves them most is the quality of mobile signal at the site, which is why a proper survey belongs at the start rather than the end.

Two points on reading that table. First, almost nothing in the first three rows is wasted when fibre arrives: a multi–WAN firewall or SD–WAN appliance simply gains a faster primary link, and the mobile circuit becomes the backup it always should have been. Resilience spent now is not a bet against the rollout; it is equipment the business needs either way. Second, the leased–line row is the only route that removes dependence on the programme altogether, and for most small rural businesses it will not be economic — which is precisely why the interim approaches above matter so much.

Two ways to handle a fibre date you cannot control

The difference between those columns is not how much is spent; it is whether the business’s continuity depends on someone else’s build programme. Every item on the right–hand side can be done today, at the address the business currently occupies, on the infrastructure that is currently there — and every one of them keeps its value after the fibre arrives.

Why this is a planning question, not just a broadband one

It is easy to file Project Gigabit under infrastructure news — something that happens to a region rather than something a business decides. That framing is understandable and it is the reason so many rural businesses end up with networks designed around an event rather than around their own operations. The more accurate framing is that connectivity is now the single dependency underneath almost everything an SME does: telephony after the copper switch–off, card payments, Microsoft 365 or Google Workspace, line–of–business applications hosted in the cloud, backups, remote access and security updates. A business that has moved all of that to the cloud has also moved its continuity risk onto its broadband line, whether or not anyone noticed it happening.

Seen that way, a fibre rollout date is an input to a technology roadmap, alongside contract renewals, hardware refresh cycles and the copper switch–off. It belongs in the same plan as those things, with the same discipline: what is the dependency, what is its current state, what is the realistic window for change, and what is the fallback if the window slips. That is the kind of question a virtual CIO or a strategic IT partner asks routinely, and it is the question that converts this month’s BDUK update from a news item into a decision.

It also shapes procurement. A business that expects fibre within twelve months should avoid signing a three–year term on its current broadband product, and should buy routing and firewall equipment that will handle a gigabit primary link when it arrives. A business whose contract is in single digits should do the opposite of waiting: invest in a resilient interim design now, on the assumption that it will be in service for some years, and treat any earlier fibre arrival as a bonus. Both are rational. What is not rational is the common middle position — doing nothing because fibre is “coming”, with no idea from which contract, which supplier or how fast that contract is actually moving.

The story at a glance

This story connects to several we have covered in the past week. The danger of having only one path to fail over to was demonstrated at national scale by Google Cloud’s UK outage, which knocked out banks and courts because there is only one UK region to fail over to — the same single–path logic, applied to a data centre rather than a rural broadband line. A business whose backups run over a slow upload link should also read ConnectWise suspending its cloud backup after M365 credential theft, which is a reminder that backup resilience depends on both the provider and the path to it. And for businesses on thin connections where large updates are already a strain, our analysis of security patches breaking ten times more often and the “patch or go offline” warning over critical PHP flaws both show why the ability to download, test and roll back updates promptly is itself a connectivity question.

Frequently asked questions

Cloudswitched

London-based managed IT services provider offering support, cloud solutions and cybersecurity for SMEs.