Investing in the invisible
HS2 shows why UK investment decisions need updating. We're connecting the world using old tools when what matters is signal strength, not just speed.
Before it was partially scrapped, HS2 was the largest infrastructure gamble in British history—an eye-watering £100 billion bet that faster meant better. Yet even as costs spiralled and the scope shrank, few paused to ask the deeper question: were we trying to solve the right problem?
The underlying assumption was simple: faster trains meant faster business. That was once true. But today, when the majority of business communication happens over screens, the need for speed has been overtaken by the need for signal. And on that measure, British rail utterly fails.
Anyone who's tried to join a video call between London and Birmingham, or hit a dead zone in Kent, knows the experience: dropped signals and silence. This isn't about rural tunnels. Dead zones exist in and around major cities in the United Kingdom. It's systemic.
The cost is not just personal frustration. One estimate suggests poor connectivity on trains costs the UK economy around £376 million annually in lost productivity. Meanwhile, it would cost just £200 million to fix it. That's a 50:1 return on investment, starkly more efficient than HS2, whose projected benefit-per-pound has fallen from £2.40 to just £1.30.
Poor connectivity on trains costs the UK economy around £376 million annually. It would cost just £200 million to fix it. That's a 50:1 return on investment, compared with HS2, whose projected benefit-per-pound has fallen to just £1.30.
Compare that with Germany. Over the past decade, Deutsche Bahn and Deutsche Telekom invested hundreds of millions of euros to ensure 99% mobile coverage across 7,800 km of mainline track. That included 800 new cell towers, window retrofitting for better signal penetration, and broadband speeds of 200 Mbps.
Switzerland ensures connectivity even through Alpine tunnels. South Korea's high-speed trains offer seamless coverage. These aren't miracles of engineering, they're deliberate choices.
Meanwhile, Amtrak in the U.S. offers a cautionary tale in the other direction. Speeds rarely exceed 15 Mbps shared across hundreds of passengers. In some cases, Wi-Fi systems have been removed entirely, and no one noticed because they never really worked to begin with.
Even planes are better connected, offering 15–20 Mbps speeds at 35,000 feet using satellite systems. If a metal tube flying at 500 mph can provide usable internet, why can't a train?
The answer lies in how we value infrastructure. Physical construction is visible, photographable, politically saleable. New lines, tunnels, and stations provide ribbon-cutting moments. Digital infrastructure does not.
I've seen this dynamic before. In 2006, I was stationed in Afghanistan when the then Defence Secretary, John Reid visited. He was shown newly rebuilt roads and bridges. But what interested him most was the scaffolding, the promise of what was still to come.
That instinct is deeply human. The familiar is ordinary and the invisible is unseen. What's being built, what will be, that's what we care about.
Just think about those who built global achievements. Is Sir Tim Berners-Lee or even Johannes Gutenberg seen in the same league as Isambard Kingdom Brunel or Ferdinand de Lesseps? They all connected the world and had to invent much more than one idea, but the wonders of the world are physical and visible.
That’s what makes technology investment so hard: you can't see the scaffolding. Those early backers of businesses that change the world, and I've made a few angel investments in the past, can't watch the building grow, or easily measure the progress; they're betting on a shift, an energy, and a team. That's hard to calibrate for structured societies, only outsiders are willing to take longshots like that.
There's a reason Silicon Valley emerged from the Californian counter-culture; those with established wealth, in New York or London, had a lot to lose if the bets went wrong, just like politicians today. Better to spend on steel and tunnels. But in the digital economy, it is precisely these hidden systems—broadband, mobile signal, secure networks—that power modern productivity.
We are now paying the price of outdated assumptions. For decades, cutting journey times was the proxy for national efficiency. But that logic no longer holds. When a few minutes saved in transit comes at the cost of hours without communication, the maths fails. We are using 20th-century tools to fix 21st-century problems.
Fixing connectivity is not just a technical upgrade; it's a cultural shift. Rail infrastructure sits between government departments, train operators, and telecoms firms. No one owns the full problem—and no one is fully accountable.
But difficult isn't impossible. Countries with clearer leadership have shown it can be done.
There is no shortage of options: trackside 5G delivers 1–3 Gbps; millimetre-wave systems offer more than 2 Gbps at high speeds; Starlink's rail antennas deliver 200–400 Mbps with low latency; onboard systems can combine multiple signals and distribute them effectively; leaky feeder cables maintain signal even in long tunnels.
These technologies exist today. The challenge is not inventing them—it's aligning policy, investment, and standards to deploy them.
Rail in the UK is cleaner than cars, faster than road travel, and more convenient than flying. Its environmental and social benefits are considerable. But its promise is being squandered by a refusal to see the invisible and the connections it brings.
This isn't about money. Nor even about trains. It's about whether Britain can evolve beyond old assumptions about what infrastructure means, and whether we're ready to connect our physical and digital worlds before our competitors leave us behind.
The technologies exist. The economic case is proven. The only question is whether we have the vision to act on it.