Coverage Is Only the Beginning

Starlink has accomplished something extraordinary: it can bring broadband service to places where building a terrestrial network may be slow, difficult, or uneconomic. A home on a rural road, a ship at sea, a commercial aircraft, or a facility far from existing infrastructure can connect through satellites hundreds of miles above the Earth. That reach is the defining strength of the system, but coverage tells us only where a signal can go. It does not tell us how many customers the network can serve in a particular place, how much data they can use, or what quality of service they will receive during the busiest hour.

Capacity Is a Local Question

Every communications network operates with finite resources, including spectrum, power, antennas, beams, and total throughput. Adding satellites and improving their performance can increase Starlink's capacity, and SpaceX's next generation of satellites represents a substantial engineering advance. The meaningful question, however, is not simply how many satellites are in orbit or how much capacity the constellation has worldwide. The question is how much usable capacity can be concentrated over a specific location when demand is at its highest, because a network with enormous global reach can still encounter local limits in heavily used markets.

The Architecture Defines the Opportunity

Starlink does not need to provide unlimited capacity everywhere to become an exceptional business. Its architecture is especially valuable in rural communities, aviation, maritime service, remote industry, military operations, and disaster recovery, where extending physical infrastructure may be impractical or take years. In those markets, the ability to reach customers quickly and reliably solves an important problem that conventional deployment methods may leave unanswered. The mistake is not believing that Starlink has enormous commercial potential; it is assuming that success in the markets best suited to its architecture means the same economics will apply in every market.

Begin With the Machine

Subscriber projections, revenue estimates, margins, and valuation multiples are outputs of a financial model, not the place where the analysis should begin. Before relying on those forecasts, I want to know where the customers are located, what service they will buy, how much data they will consume, how often spectrum can be reused, and what additional investment will be required to support that demand. Those answers begin with the physical system and the constraints that govern its performance. Capital can finance extraordinary engineering, but it cannot negotiate with physics, which is why a credible valuation of Starlink must begin with what the network can actually deliver.