Picture this: you board a train at London Paddington, heading west toward Bristol. You have a laptop open, a Teams call in thirty minutes, and a phone showing a healthy-looking 5G icon. Within ten minutes of departure, you are staring at a spinning wheel. The signal icon flickers from 5G to 4G, then to ‘H’ — the industry’s euphemism for HSPA, a technology that was already outdated a decade ago — before dropping to ‘E’ for EDGE, a standard from the early 2000s. Your video call is dead before it started. Your email refuses to send. You are, in the sixth-largest economy in the world, functionally offline.
This is not a fringe experience confined to the Scottish Highlands or remote Welsh valleys. This is the daily reality for millions of commuters, business travellers, and ordinary citizens across England’s busiest rail corridors and urban centres. And the question that demands an answer is not simply “why is it this bad?” but “why has the regulator, Ofcom, allowed it to remain this bad for so long?”
The Scale of the Problem
The United Kingdom’s mobile network performance is, by almost every credible international benchmark, embarrassing. In a July 2025 study by network monitoring firm MedUX, which examined data from millions of users across more than forty countries, the UK was found to have 5G availability of below 35 per cent. That means roughly only one in three mobile data sessions in the UK utilised a 5G connection, even when the user had a 5G-capable handset and was in an area with supposed 5G coverage. Countries such as Greece, the Czech Republic, Denmark, Luxembourg and the Netherlands all achieved 5G availability exceeding 65 per cent.
The picture grows darker when you examine download speeds. The UK’s average 5G download speed sits below 51Mbps according to MedUX, with upload speeds under 14Mbps. Packet loss exceeds 0.6 per cent — double that of Switzerland. For gamers, the UK recorded the highest average jitter in Europe at 33.48 milliseconds. For video streamers, the country has among the highest stalling ratios on the continent at over 0.4 per cent.
An earlier study by the Social Market Foundation, a cross-party think tank, ranked the UK dead last among 15 developing and advanced international markets for both 5G availability and download speeds. UK users received 5G coverage just 10 per cent of the time, compared with 43 per cent in India and 38 per cent in South Korea. Average UK 5G download speeds of 118Mbps were dwarfed by South Korea’s 437Mbps. India — a developing nation with vastly greater geographical complexity — managed 281Mbps.
Even London, the financial capital of Europe and a city that positions itself as a global tech hub, performs abysmally. In early 2025, Ookla’s Speedtest data found that London trailed other UK cities in 5G network consistency, median download speeds, and upload speeds. This makes the UK unique in Western Europe: in no other major European nation does the capital city sit at the bottom of its own domestic rankings. On the global stage, London ranks as one of the slowest European cities for 5G speeds — approximately 75 per cent slower than Lisbon, according to Opensignal analysis.
Ookla’s Speedtest Global Index paints the broadest picture: as of its latest rankings, the UK places 55th out of 104 countries for median mobile speeds at around 70Mbps. It sits behind not only Denmark (198Mbps), the Netherlands (171Mbps) and France (135Mbps), but also behind Brazil (210Mbps), China (171Mbps), India (152Mbps) and Vietnam (134Mbps). The UK does not merely fall behind its peers — it falls behind the global median of 91Mbps.
On the Rails: A Connectivity Black Hole
Nowhere is the UK’s mobile connectivity failure more acutely felt than on its railway network. Britain’s rail system stretches nearly 10,000 miles, carrying almost five million journeys daily. Yet in 2025, a study by Streetwave testing mobile data performance between London and Sheffield found that across all four major operators, essential coverage — defined as the percentage of the route with a stable, usable signal — was alarmingly low. For over two‑thirds of the journey, none of the major networks could sustain a stable data connection.
Median download speeds recorded on that route were barely enough to load a web page, let alone support video calls, real-time applications, or streaming. Between rural segments, tunnels, and cuttings, the signal drops away entirely, leaving long digital dead zones where passengers are cut off completely. As Clarus Networks described it:
This isn’t an occasional dropout; it’s a structural connectivity gap that affects every service, every day.
For commuters who rely on train journeys as productive work time, this is a direct economic loss. Business travellers cannot join video calls. Commuters cannot check emails or access live journey updates. Passengers increasingly reliant on digital ticketing and mobile payments find themselves stranded without functionality. And for the rail operators themselves, the consequences run deeper: modern trains depend on connectivity for real-time diagnostics, CCTV backhaul, crew communications, and predictive maintenance. When networks fail mid‑route, data gaps emerge that can compromise safety, delay repairs, and increase operating costs.
The UK government’s response has been “Project Reach,” a partnership signed in June 2025 between Network Rail, Neos Networks, and Freshwave, which will deploy fibre-optic cable along mainline corridors and install mobile infrastructure in 57 tunnels covering almost 50 kilometres. The first installations are not expected until 2026, with full rollout by 2028. This is welcome, but it comes absurdly late. In December 2017, the then-government pledged “uninterrupted” Wi‑Fi and mobile speeds of up to 1Gbps on all mainline train routes by 2025. That promise was quietly abandoned. The fact that in 2026 we are only just beginning to address tunnel blackspots on routes as critical as the East Coast and West Coast Main Lines is a damning indictment of a decade of inaction.
Underground and Left Behind
London’s Underground network — the world’s oldest metro system, carrying over five million journeys per day — only began receiving 4G and 5G mobile coverage in earnest from 2022 onwards. By mid‑2025, coverage extends to sections of the Central, Northern, Jubilee, Victoria, and Piccadilly lines, along with the full Elizabeth line. But the rollout is far from complete. Many stations and tunnelled sections remain without coverage. The full project is not expected to be finished until 2026, with some outlying sections potentially slipping into 2027.
To put this in context: Seoul’s metro has had comprehensive mobile coverage for over a decade. Tokyo’s underground system has provided uninterrupted 4G since the mid‑2010s. Moscow’s metro has offered free Wi‑Fi across its entire network since 2014. Even cities like Bucharest, Istanbul, and Delhi have beaten London to full underground connectivity. The idea that the capital of the world’s sixth‑largest economy has passengers regularly losing signal between stations in 2026 should be a source of national embarrassment.
Why Is It This Bad? The Structural Failures
The Huawei Ban: A Necessary Decision with Devastating Consequences
In July 2020, the UK government banned Chinese telecoms giant Huawei from supplying new equipment for 5G networks and ordered all existing Huawei 5G equipment to be removed by the end of 2027. The decision, based on National Cyber Security Centre analysis following US sanctions on Huawei’s supply chain, was arguably the right call on security grounds. But its impact on network rollout has been devastating.
The government itself acknowledged at the time that the ban would delay the UK’s 5G rollout by up to three years and cost the industry up to GBP2bn (USD2.7bn). In practice, the damage may have been even worse. Three UK reported that nearly all of its GBP450mn (USD609mn) capital expenditure in one recent year went toward replacing Huawei equipment or the Shared Rural Network — not toward expanding coverage or improving service quality. BT estimated its total Huawei swap-out cost at GBP500mn (USD677mn). Vodafone put its costs in the low single-digit billions. These are vast sums diverted from network improvement into what is essentially treading water.
Before the ban, Huawei was a key supplier that kept costs competitive and rollout moving quickly. Without it, operators scrambled to replace equipment from Nokia and Ericsson at higher cost and slower pace. Rural areas, where the business case for investment is already marginal, bore the heaviest burden. The billion-pound Shared Rural Network, meant to close coverage gaps, instead saw operators diverting resources to rip out Huawei kit rather than build new coverage.
Spectrum Mismanagement and the Non‑Standalone 5G Problem
The UK’s 5G spectrum was auctioned in a piecemeal fashion: mid-band frequencies around 3.4GHz were sold in 2018, frequencies in the 3.6‑3.8GHz range in 2020, and low-band 700MHz spectrum not until 2021. This fragmented approach meant operators could not secure the spectrum they needed when they needed it, discouraging the investment required to build genuinely capable 5G infrastructure.
Most critically, the UK’s 5G networks remain overwhelmingly “Non-Standalone” (NSA), meaning they are bolted onto existing 4G infrastructure rather than operating independently. This dramatically limits 5G’s potential. True Standalone 5G (SA) delivers the speed, latency, and capacity improvements that 5G was designed to provide. Without it, UK consumers are essentially experiencing a moderately faster version of 4G with a 5G icon on their phone. The government has set a target of nationwide SA 5G in all populated areas by 2030, but progress toward that goal has been glacial.
Planning Barriers and Underinvestment
Building mobile masts in the UK remains a tortuously slow process. Complex planning regulations, local opposition driven by unfounded health concerns or aesthetic objections, and protracted consultation periods mean that deploying new infrastructure takes far longer than it should. The government has reformed permitted development rights to allow taller and wider masts without full planning permission, but public consultation requirements persist, and a vocal minority can delay or block installations entirely.
The Shared Rural Network, originally budgeted at GBP1bn (USD1.35bn) with 260 new masts, has been scaled back to just 60 new masts. The original ambition was already modest by international standards; the scaled-back version is barely scratching the surface.
The Ofcom Question: Where Is the Regulator?
This brings us to the central question: what has Ofcom, the UK’s comms regulator, been doing while the UK’s mobile networks have fallen so catastrophically behind the developed world?
Ofcom’s principal statutory duty under the Communications Act 2003 is “to further the interests of citizens in relation to communications matters.” It is difficult to see how presiding over one of Europe’s worst-performing mobile networks fulfils that duty. The regulator’s failings can be grouped into three categories.
1. Reliance on Operator-Provided Data
Until mid‑2025, Ofcom’s mobile coverage data was based primarily on predictions provided by the mobile network operators themselves. These predictions were generated by computer models simulating how signals travel from masts, accounting for obstructions like hills and buildings. The data was processed into a 100-metre-by-100-metre grid. Ofcom itself has acknowledged that this data “may not always match people’s real-world experience at a very local level.”
Let that sink in: the regulator charged with ensuring UK citizens receive adequate mobile coverage was, for years, relying on the operators it regulates to tell it whether their own coverage was adequate. This is akin to asking a restaurant to grade its own hygiene inspection. It is a fundamental abdication of regulatory responsibility.
The results were predictable. As the UK Government’s own minister stated in Parliament in March 2025: “Ofcom’s reporting of mobile coverage is rather over-optimistic and does not reflect people’s lived experience.” The minister went further, noting that “the connectivity that people think they are getting from Ofcom is simply not what they are actually getting. Their phone looks as though it has lots of bars and is saying 4G, but they cannot even download an app to park their car.”
Independent testing tells a starkly different story from Ofcom’s published figures. Streetwave, a commercial mobile mapping company, tested coverage across 113 local council areas and found that EE offers acceptable coverage in 69 per cent of the UK, Vodafone 61 per cent, O2 50 per cent, and Three just 38 per cent. Compare these figures to Ofcom’s headline claim of 96 per cent 4G coverage from at least one operator, and the gulf between reported coverage and lived experience becomes clear. As the Public Accounts Committee urged in 2024, the government must “take urgent action to ensure that it has meaningful data on mobile coverage that reflects people’s actual experience.”
2. Inadequate Independent Testing
A regulator with teeth would conduct its own rigorous, nationwide, systematic testing of mobile network performance. Ofcom’s own methodology document admits that its testing consists of “periodic drive testing” across “a sample set of different locations.” This is done primarily on roads and does not test network capacity. The regulator acknowledges that these tests show predictions are “on average reasonable, but subject to a significant prediction error at a local level.”
In June 2025, Ofcom launched an updated “Map Your Mobile” coverage checker that incorporates crowdsourced data from Opensignal alongside the operators’ predictions. This is a step forward, but it remains fundamentally limited. Crowdsourced data depends on where volunteer users happen to be; it cannot replace systematic, controlled measurement across the full UK geography, including inside buildings, on trains, and in areas of known poor coverage.
Other regulators take a more hands‑on approach. Germany’s Bundesnetzagentur conducts extensive drive testing and publishes detailed, independently verified coverage maps. It has also imposed financial penalties on operators for missing coverage obligations. In the United States, the Federal Communications Commission conducts its own broadband mapping initiatives and has fined carriers for misrepresenting coverage. Ofcom’s reliance on operator-generated data and limited sampling represents a level of regulatory passivity that would be unacceptable in most comparable economies.
3. Weak Enforcement and Toothless Obligations
Ofcom’s coverage obligations for mobile operators are astonishingly unambitious. For 4G, the definition of coverage is based on the minimum signal strength required to deliver a 98 per cent probability of completing a 90‑second voice call. For data, the threshold is a 95 per cent chance of achieving a download speed of at least 2Mbps. In 2026, when consumers expect to stream high-definition video, participate in video conferences, and use data-intensive applications, a 2Mbps threshold is an insult.
Even when operators fail to meet these already‑low thresholds, Ofcom’s enforcement has been remarkably gentle. When Three UK initially failed to meet its Shared Rural Network interim coverage obligation of 88 per cent by June 2024, and also fell short in Scotland, Ofcom subsequently confirmed Three met the targets by August 2025 and announced it “would not be taking enforcement action.” Three operators — Vodafone, Virgin Media O2, and Three — had all requested an 18‑month extension to their SRN obligations, which the government rejected. But the consequence of their tardiness was… nothing.
The government and Ofcom also lack the legal power to force operators to build masts in specific locations. Coverage obligations are geographic averages, meaning operators can meet targets by covering large, open areas while leaving dense urban pockets, commuter corridors, and railway routes effectively unserved. The result is that 96 per cent coverage on paper coexists with routine service failures in places where people actually need connectivity.
The Edge of Absurdity: Why Are We Still Falling Back to 2G?
Perhaps nothing captures the state of UK mobile connectivity more vividly than the persistent appearance of the ‘E’ symbol on smartphones — denoting EDGE (Enhanced Data Rates for GSM Evolution), a 2G technology that dates from the early 2000s. For a connection to fall back to EDGE in 2026 means you are using infrastructure designed for an era when a polyphonic ringtone was cutting-edge technology.
This happens because UK mobile networks deploy spectrum inefficiently. When a 4G or 5G signal weakens, the phone attempts to latch onto the next available technology. In areas of poor coverage — which, as we have established, encompasses a significant proportion of the UK’s geography — the only available fallback may be 2G or 3G. Since operators are actively shutting down 3G networks (with full Ofcom support) to repurpose the spectrum for 4G and 5G, the perverse result is that the fallback in areas of poor coverage is increasingly not the moderately usable 3G, but the functionally useless EDGE.
In other countries, this problem has been addressed through denser site deployment, better spectrum utilisation, and a commitment to ensuring that the next-generation network actually covers the footprint before the previous generation is switched off. In the UK, we are instead creating an expanding patchwork of connectivity deserts where phones display impressive-sounding technology labels while delivering speeds that would embarrass a 2003 dial‑up modem.
A Challenge to Ofcom: Five Demands
Ofcom has had years to address the UK’s mobile connectivity crisis. It has instead pursued an approach characterised by deference to operators, dependence on self-reported data, and a reluctance to impose meaningful consequences for poor service. It is time for the regulator to be held to account. I have five demands.
First: Conduct independent, systematic network testing nationwide. Ofcom must stop outsourcing its core regulatory function to the companies it regulates. It should commission or conduct comprehensive drive-testing, walk-testing, and in-building testing across urban, suburban, rural, and transport environments. Results must be published transparently and used as the primary basis for assessing coverage obligations — not operator‑generated predictions.
Second: Set coverage obligations that reflect how people actually use mobile networks. A 2Mbps download threshold is a relic. Minimum performance standards should be set at levels that support modern use cases: video calling, cloud‑based applications, streaming, and real‑time navigation. Coverage obligations must also specifically address transport corridors, including mainline rail routes, motorways, and underground networks, rather than treating them as afterthoughts.
Third: Impose financial penalties for failure to deliver. Coverage obligations without meaningful enforcement are performance theatre. When operators miss their targets, there must be substantial financial consequences. Spectrum licence conditions should be tightened to include quality-of-service requirements, not merely geographic coverage benchmarks based on signal prediction models.
Fourth: Publish real‑time, granular quality data. Consumers deserve to know not just whether a signal exists in their area, but how that signal actually performs. Ofcom should mandate that operators publish real‑time network performance data, including average speeds, latency, and reliability, at a granular geographic level. This data should be independently verified.
Fifth: Provide an annual public account of the UK’s international standing. Ofcom should be required to publish, as part of its Connected Nations report, a frank assessment of how the UK’s mobile performance compares internationally. Not as a footnote or a caveat, but as a headline finding. If the UK ranks 55th globally for mobile speeds, the regulator should say so plainly and explain what it intends to do about it.
Conclusion: This Is a Choice, Not an Inevitability
The UK’s mobile connectivity crisis is not a natural disaster. It is the result of policy choices, regulatory timidity, and chronic underinvestment. The Huawei ban imposed genuine costs, but it was a security decision that other countries — including Australia, which banned Huawei even earlier — have managed without falling as far behind. The planning system is cumbersome, but countries with equally complex regulatory environments have built better networks. The market is competitive, but competition without effective regulation merely produces a race to the bottom, where operators compete on price while collectively underinvesting in infrastructure.
Ofcom sits at the centre of this failure. A regulator that relies on the companies it oversees to report their own performance, that sets coverage thresholds a generation behind consumer expectations, that declines to penalise operators who miss their targets, and that presents coverage statistics wildly divorced from lived experience is not fulfilling its statutory duty. It is providing cover for an industry that charges premium prices for a substandard service.
The economic costs are real. The Social Market Foundation estimates that widespread 5G availability could add GBP159bn (USD215bn) in economic benefit by 2035 through operational efficiencies and innovation. Every year of delay is lost productivity, lost growth, and lost competitive advantage against nations that are building the digital infrastructure of the future while the UK remains stuck, watching a spinning wheel on a train platform, waiting for a signal that may never come.
The UK public deserves better than this. It is time for Ofcom to stop being a passive observer of Britain’s mobile decline and start acting like the regulator the country needs.
