Satellite is having a moment. But are we asking the right question?

This week, NTT DOCOMO announced that its Starlink Direct service has surpassed five million users in just over two months since launch. On the surface, it is another impressive milestone for direct-to-device satellite communications. It demonstrates the speed at which consumers are adopting services that extend connectivity beyond the reach of traditional terrestrial networks. Yet, for me, the real significance of the announcement lies elsewhere. It is not about five million users. It is about what those five million users represent.

For decades, the telecommunications industry has pursued a remarkably simple ambition: ubiquitous connectivity. Every new generation of mobile technology has promised broader coverage, greater capacity and faster speeds. Operators have invested hundreds of billions of dollars building towers, acquiring spectrum and extending fibre networks, steadily reducing the number of places where a customer cannot connect. But the final few per cent has always proved the hardest. Whether it is a mountain range, a remote island, an offshore wind farm or simply a rural community where the economics do not stack up, there are still places where terrestrial infrastructure struggles to reach.

Historically satellite communications have played a niche role within the telecommunications industry. They largely served shipping, aviation, defence, disaster recovery and specialist users carrying dedicated devices. And they were expensive to use.

Today, huge investment in satellite connectivity is changing the economic calculation and that distinction is disappearing. Increasingly, satellite is becoming another access technology within the mobile ecosystem itself, allowing the smartphone already in your pocket to connect when the nearest cell tower is simply too far away.

Over the past few years we have seen operators across the world announce partnerships, commercial launches and trials exploring direct-to-device satellite services. T-Mobile has partnered with Starlink in the United States, Vodafone has successfully demonstrated satellite-to-smartphone connectivity in Europe, AST SpaceMobile continues to develop its direct-to-device capabilities with operators around the world, and satellite is becoming a strategic priority for many communications service providers. Individually, these announcements have generated plenty of headlines. Collectively, they suggest something much more significant: satellite is moving from innovation to implementation.

Architecting customer experience

The obvious conclusion is that satellite will finally solve the industry's remaining coverage gaps. But the story isn’t just about coverage. It is also about architecture.

Operators are now seriously discussing networks that extend beyond the Earth's surface as a natural extension of their existing infrastructure. Rather than thinking about terrestrial and non-terrestrial networks separately, we are beginning to think about them as one integrated platform. If that vision becomes reality, then the customer experience changes fundamentally. Consumers should not need to know whether they are connected through a macro site, a small cell, Wi-Fi or a satellite orbiting hundreds of kilometres above the Earth. They should simply remain connected.

That seemingly simple ambition raises profound questions for our industry. How should operators orchestrate traffic across terrestrial and non-terrestrial networks? How should assurance, service management and automation work when part of the network is in low Earth orbit? How do we expose consistent capabilities to developers regardless of how a customer happens to be connected? Most importantly, how do we avoid creating another isolated technology silo just as the industry has spent years trying to simplify its architecture?

This is where interoperability becomes every bit as important as the satellite technology itself. If non-terrestrial networks become another layer of future telecommunications infrastructure, they cannot rely on proprietary interfaces and bespoke operational models. They need to integrate seamlessly into the digital architectures operators are already building. Open APIs, common information models and composable platforms become just as important in space as they are on the ground.

That is precisely why initiatives such as TM Forum's Satellite Open Digital Architecture (SODA) project matter. The challenge is not simply enabling satellites to connect to mobile networks. It is ensuring they become part of the same interoperable ecosystem, allowing operators to innovate without creating new complexity.

We are already seeing this vision come to life through TM Forum's Catalyst programme. Projects such as SATCOM with an Edge – Phase III have demonstrated how satellite, 5G and SD-WAN can be orchestrated as a single service, enabling operators to order, assure and manage seamless connectivity across terrestrial and non-terrestrial networks using TM Forum Open APIs. Rather than treating satellite as a standalone technology, the project showed how it can become another integrated component of a modern, composable telecom architecture. More recently, the Hyperconnected Intelligent Network of Trust (HINT) – Phase III Catalyst has expanded that vision even further. Rather than viewing satellite purely as a solution for extending coverage, the project explores how converged satellite-mobile connectivity, AI and intelligent orchestration can enable entirely new enterprise services, including real-time supply chain visibility, predictive logistics and trusted digital ecosystems. It is a powerful reminder that the future of satellite is unlikely to be defined by connectivity alone. Its greatest value may lie in the new digital services and business models it enables. Taken together, these initiatives highlight something important. The industry's ambition should not simply be to connect satellites to mobile networks. It should be to ensure satellite becomes another interoperable, programmable and intelligent layer within tomorrow's digital telecom platform.

Building an intelligent platform

It is also worth recognising that we are still at the beginning of this journey. Today's direct-to-device services remain relatively limited, and terrestrial networks will continue to carry the overwhelming majority of mobile traffic for many years. Their capacity, latency and economics in densely populated areas simply cannot be matched by today's satellite technologies. The commercial models are still evolving, regulation will continue to mature and many technical challenges remain.

However, the conversation has undeniably changed.

For years we have debated whether the next competitive advantage would come from 5G, Open RAN, cloud-native networks or AI-native operations. Increasingly, another question is emerging. What happens when every operator has the ability to extend its network into space? As the industry continues its pursuit of ubiquitous connectivity, perhaps the question is no longer whether satellite has a role to play. It clearly does. The more interesting question is how far that role will evolve over the next decade. Terrestrial networks are unlikely to disappear. Just as fibre did not replace wireless, and Wi-Fi did not replace cellular, satellite is unlikely to replace terrestrial infrastructure outright. Instead, it may become another layer within an increasingly intelligent, hybrid communications platform, with each technology complementing the others according to geography, economics and customer need.

But technology has a habit of reshaping industries in ways we rarely predict. As launch costs continue to fall, satellite technology matures and non-terrestrial networks become more deeply integrated into operator architectures, it is no longer unreasonable to ask whether future mobile networks will be designed with space as a fundamental layer rather than simply an extension of terrestrial infrastructure.

Perhaps, then, the question we should be asking isn't whether satellite will replace terrestrial networks. It's whether, in ten years' time, we'll still think of them as separate networks at all.