Satellite is becoming a standard tool, not a niche solution
Across Europe, a quiet infrastructure shift is underway — not a revolution, but a filling-in of the gaps that geography, economics, and political will have left behind. Starlink's median download speeds across 27 European markets rose 45 percent year-over-year in early 2026, reaching 165 megabits per second, with the strongest adoption concentrated precisely where mountains, islands, and dispersed settlements have made terrestrial networks slow to arrive. The data does not tell a story of satellite displacing fiber, but of satellite finding its natural home in the places fiber has not yet reached — and perhaps never will on purely commercial terms. What is emerging is a hybrid model, where sovereign ambition, private innovation, and the stubborn realities of terrain are quietly reshaping how Europe thinks about universal connectivity.
- Starlink's 45% speed surge across Europe is real, but the headline masks a more precise truth: the gains matter most in Bulgaria, Greece, and Croatia, where satellite fills a vacuum that terrestrial networks have left open for years.
- Bulgaria's 8% Starlink sample share — the highest in Europe — comes with a warning: concentrated demand has pushed median speeds down to just 61 Mbps, the weakest on the continent, showing that satellite capacity is not yet infinite.
- In fiber-rich Denmark, the Netherlands, and Romania, Starlink barely registers below 1% sample share, proof that where fast, affordable terrestrial networks exist, satellite finds no foothold.
- Fixed networks still hold decisive advantages in latency and upload speeds across virtually every market, keeping satellite in a supporting role rather than a leading one for most use cases.
- The UK, Amazon Leo, and Europe's own IRIS² program signal that satellite broadband is graduating from a rural curiosity into a formal layer of national connectivity strategy — integrated into subsidy design and resilience planning.
In the first quarter of 2026, Starlink's median download speeds across 27 European markets climbed from 114 to nearly 166 megabits per second — a 45 percent year-over-year gain driven by SpaceX's expanded constellation of more than 9,000 active satellites. But the more revealing story lies not in the aggregate number, it lies in where the service is actually being used.
The highest adoption rates cluster in countries where geography has made universal broadband slow and expensive to deliver. Bulgaria led all markets with 8 percent of Speedtest samples coming from Starlink, followed by Greece and Croatia at 6 percent each — nations of islands, mountains, and dispersed settlements where national median speeds lag far behind what satellite can now offer. Latvia illustrated the gap starkly: Starlink's 232 Mbps median outpaced the national fixed broadband median by 85 megabits. Even Ireland, despite one of the world's most ambitious rural fiber programs covering over 560,000 premises, ranked Starlink fourth in adoption — a signal that even well-funded rollouts leave edges, and that some households treat satellite as resilience against routine winter line failures.
Elsewhere, satellite barely registers. In Denmark, the Netherlands, Finland, and Romania, Starlink's sample share fell below 1 percent. Romania is particularly instructive: a large rural population, yet fiber-first operators like DIGI have already delivered gigabit service at low cost across the country, leaving satellite with no meaningful gap to fill. Spain, Europe's fiber leader with a national median of 278 Mbps, treats satellite as a precision instrument for its remaining uncovered areas — not a mass-market alternative.
The performance picture carries its own nuances. Starlink outpaced average fixed network speeds in 11 of 27 markets on median downloads, but fixed networks held advantages in latency and upload speeds in nearly every country. Bulgaria offered the clearest caution: highest adoption, lowest Starlink speeds, and the only market where performance actually declined year-over-year — a reminder that concentrated demand can strain satellite capacity in ways that may persist as uptake grows.
What the data ultimately describes is complementarity rather than disruption. Starlink is becoming essential infrastructure for farms, islands, mountain communities, and the final fraction of premises that terrestrial networks have not reached. The UK has formalized this logic through Project Gigabit, partnering with BT Group to deploy Starlink specifically for its hardest-to-reach premises. Meanwhile, Eutelsat OneWeb, Amazon Leo, and Europe's sovereign IRIS² program signal that satellite broadband is broadening into a standard layer of connectivity strategy — integrated into subsidy design, resilience planning, and the long-term question of what universal access actually means in a continent where the last mile is sometimes a mountain pass.
Across 27 European countries, Starlink's download speeds climbed 45 percent in the first quarter of 2026 compared to the same period a year earlier, jumping from 114 megabits per second to nearly 166. The improvement is real and measurable, driven by SpaceX's expansion of its satellite constellation to more than 9,000 active satellites and the addition of millions of new customers worldwide. But the story the speed gains tell is not one of wholesale replacement of Europe's terrestrial networks. Instead, it reveals where those networks still fall short—and where satellite broadband is quietly becoming essential infrastructure.
The pattern is clearest in the places where fiber, cable, and fixed wireless have yet to reach. In Bulgaria, Starlink accounted for 8 percent of all Speedtest samples in the quarter, the highest share across all measured markets. Greece and Croatia followed at 6 percent each. These are not wealthy, densely populated urban cores with mature broadband ecosystems. They are countries where geography—islands, mountains, dispersed settlements—has made universal high-speed coverage slow and expensive to deliver. Greece's national broadband median sits at 94 megabits per second; Starlink there hit 196. Latvia showed an even starker gap: Starlink's median of 232 megabits per second outpaced the national fixed median by 85 megabits. In Ireland, despite one of the world's largest state investments in rural broadband covering more than 560,000 premises and 1.1 million people, Starlink ranked fourth in adoption—a sign that even ambitious fiber rollouts leave gaps, and that some households value satellite as backup against the winter storms that have become routine fiber line failures.
But satellite broadband is not winning everywhere. In seven markets—Denmark, Malta, Finland, Romania, Luxembourg, the Netherlands, and Slovenia—Starlink's sample share fell below 1 percent. These are places where fiber networks are dense, speeds are fast, and prices are competitive. Romania is instructive: despite a large rural population, Starlink barely registers because fiber-first operators like DIGI have already blanketed the country with gigabit service at low cost. Spain, Europe's fiber flagship, saw national median speeds of 278 megabits per second, more than 110 megabits faster than Starlink. The country's Conectate35 program offers satellite broadband as a targeted tool for the remaining uncovered areas, treating it as a precision instrument rather than a mass-market solution.
The performance picture is mixed in other ways. Starlink was faster than the average fixed network on median download speed in 11 of the 27 markets. But fixed networks held decisive advantages in latency—the delay in data transmission—across every single country. Upload speeds favored fixed networks in 26 of 27 markets. Bulgaria offered a cautionary note: it had the highest Starlink adoption but the weakest median download speed at just 61 megabits per second, and it was the only market where speeds actually declined year-over-year, down 5 percent. The pattern suggests that concentrated demand can still strain available satellite capacity in some locations, a constraint that may persist as adoption grows.
What emerges from the data is not a story of disruption but of complementarity. Starlink is becoming a credible option in the hardest-to-serve places—farms, islands, mountain communities, seasonal homes, remote businesses, and the final few percent of premises that terrestrial networks have not yet reached. It works alongside fiber-led policy, not against it. The UK has formalized this through Project Gigabit and a BT Group agreement to use Starlink for the hardest-to-reach premises, and the country's median latency of 37 milliseconds suggests that dense ground station infrastructure can optimize performance even for satellite service.
The satellite broadband market itself is broadening. Eutelsat OneWeb operates a 600-plus satellite network focused on enterprise and government use. Amazon Leo began full-scale deployment in April 2025 and had more than 300 satellites in orbit by late April 2026, though it has not yet appeared as a residential force in European speed tests. Europe is building its own sovereign layer through IRIS², designed around secure, government-controlled connectivity. Together, these services signal that satellite broadband is moving beyond the niche of one dish on one remote home. It is becoming part of the standard toolkit for delivering connectivity to places where the economics of terrestrial networks do not work. The question now is not whether satellite will replace fiber, but how it will be integrated into Europe's long-term broadband strategy—and what role it will play in the subsidy design, resilience planning, and rural connectivity decisions that lie ahead.
Citas Notables
Starlink is strongest where the last few percent of premises are hardest to serve—islands, farms, mountains, seasonal homes, and rural businesses still waiting for planned upgrades.— Ookla Speedtest Intelligence analysis
La Conversación del Hearth Otra perspectiva de la historia
Why does Starlink show up so much more in places like Greece and Bulgaria than in Denmark or the Netherlands?
Because those countries have already solved the broadband problem with fiber. When you have fast, cheap, widely available fixed networks, most people use them. Starlink appears where the terrestrial networks have given up or haven't arrived yet—islands, mountains, places where the economics don't work for fiber companies.
But Starlink is faster than the national average in Greece. Shouldn't that make it more attractive even in places with decent fixed broadband?
Speed alone doesn't drive adoption. Latency matters enormously for real-time applications, and fixed networks beat satellite on that in every single European market. Upload speeds too. Starlink is a solution to a specific problem—coverage in hard-to-reach places—not a universal upgrade.
What happened in Bulgaria? It had the highest adoption but the slowest speeds.
That's the constraint showing itself. When too many people in one area rely on the same satellite capacity, the network gets congested. Bulgaria suggests there's a ceiling to how much demand one region can absorb before performance degrades.
Is Starlink replacing Europe's fiber plans?
No. Ireland is the clearest example—they've invested billions in fiber for rural areas, and Starlink still ranks fourth in adoption. People use it for the areas still waiting for fiber and as backup when storms knock out lines. It's complementary, not competitive.
What about Amazon Leo and these other satellite networks?
They're part of a bigger shift. Satellite is becoming a standard tool in the broadband toolkit, not a niche solution. But each service has a different purpose. OneWeb focuses on enterprise and government. Amazon Leo is still ramping up. Europe's IRIS² is about sovereign control. Starlink is the only one that's already a mass-market residential option.
So what's the real story here?
Europe's broadband landscape is becoming hybrid. Fiber gets the dense, profitable areas. Satellite handles the last few percent—the places where geography or economics make terrestrial networks impractical. The question now is how to design subsidies and resilience strategies around that reality.