SpaceX Wants 80,000 Towers. It Does Not Want to Own Them.

Next time you are stuck in traffic, look up.

Not at the sky. At the side of the road.

There is a tower there. A plain metal tower, with antennas stuck on it like shelves. You have driven past a thousand of them and never really looked at one. They are boring. They are ugly. Nobody takes a photo of a cell tower on holiday.

Last week, those boring towers became the most interesting business in America.

In the space of about two weeks, SpaceX went from a rocket and satellite company to a company that might become America's fourth phone carrier. AT&T, Verizon and T-Mobile fell. The companies that own the towers jumped. And a lot of people got the story wrong in the same way.

They said SpaceX is buying 80,000 towers.

It is not. And the difference between buying a tower and renting space on one is the whole story.

What actually happened

Let us start with the facts, because the headlines moved faster than the facts did.

On October 8, Grain Management announced a deal to sell SpaceX its entire nationwide 800 MHz spectrum portfolio. The deal still needs approval from the FCC, the regulator that controls who can use which airwaves in the US. The companies did not announce a price. Reports put it at about $8 billion.

Two weeks later, on October 22, Bloomberg reported something else. SpaceX is in talks with tower companies to get access to at least 80,000 cell sites across the US. Crown Castle was named in the report. Crown Castle's chief executive was asked about it on an earnings call and did not confirm SpaceX by name. He just said, you can assume we are talking to all parties.

That is it. Talks. Not a signed deal. Not a purchase.

The next day, the market reacted like the deal was already done. AT&T, Verizon and T-Mobile fell more than 4% before the market even opened. Crown Castle jumped about 12%. American Tower rose too.

I think the market was not reacting to what SpaceX has built. It was reacting to what SpaceX might become.

First, what is spectrum, in plain language

Your phone does not talk to a tower with a wire. It talks with radio waves.

Think of those radio waves as lanes on a highway. There are only so many lanes. The government decides who gets to use which lane, in which city, and for what. If you do not own a lane, you cannot run a phone network. It does not matter how good your satellites are.

SpaceX just bought, or is trying to buy, a very specific kind of lane: low-band spectrum at 800 MHz.

Low means low frequency. And low frequency has a simple trade-off that anyone can understand.

Low waves travel far. They go through walls, trees and windows better than high waves. Think of bass from a neighbour's speaker. You do not hear the high notes through the wall. You hear the bass. It is the same idea.

High waves carry more data and go faster, but they die quickly. A wall can stop them.

That is why 800 MHz matters so much. It is the coverage lane. It is the lane that can reach inside a building, down a rural road, and across a wide area without needing a tower on every corner.

SpaceX already has something else: mid-band spectrum around 2 GHz, from its earlier deal with EchoStar. That is for capacity, for carrying a lot of data. The 800 MHz lane is for coverage, for making sure the signal actually gets to you when you are indoors.

And here is a small detail that matters: most phones people already own can use the 800 MHz band. SpaceX does not need everyone to buy a special SpaceX phone to make this work.

So why does a satellite company need towers at all?

This is the question I kept coming back to. If SpaceX has thousands of satellites in space, why is it now talking to tower companies on the ground?

Because satellites have a weakness, and it is a very human weakness. They cannot get through walls very well.

A satellite is hundreds of kilometres away. By the time its signal reaches your phone, it is weak. Out in the open, in a field or on a highway or on a boat, that is fine. Inside an apartment building in Fremont, or inside a shopping mall, or in a stadium with 50,000 other phones, it is not fine.

A tower is the opposite. It is close. It is strong. It can handle a lot of phones in a small area. But it only covers a few miles.

So SpaceX is not choosing satellites instead of towers. It is trying to build both, and let your phone switch without you noticing.

Satellite when there is no tower for miles. Tower when you are in a city, or indoors, or somewhere crowded.

That is what people mean when they say hybrid network. It is not a fancy word. It just means ground plus space, together.

Without towers, Starlink Mobile stays a backup service. A good backup service, for emergencies and dead zones, but still a backup.

With 80,000 sites, it starts to look like a real carrier.

80,000 sites is not 80,000 towers

This is where most people get confused, and I do not blame them. The words sound the same.

A tower is the steel structure. A site is a rented spot on that structure, or on a rooftop, or on a water tank, where one company puts its own antennas, radios, power and fibre connection.

Three or four carriers can all be on the same tower. They each pay rent to the tower owner. The tower owner does not care which carrier wins. It collects rent from all of them.

That is why the tower companies jumped when the carriers fell.

Crown Castle owns about 40,000 US towers. American Tower owns about 42,000 US sites. A private company called Vertical Bridge owns about 18,000. Add them up and you start to see why SpaceX needs to talk to more than one company to reach 80,000 sites.

For scale, there were about 158,500 purpose-built cell towers operating in the US last year. So 80,000 sites is roughly half of that. It is not a test. It is not a small experiment in one state. It is national scale.

But it is still renting, not owning. And renting is faster than building. One tower executive called the existing tower network an industry cheat code, because building that many towers from scratch would take years.

SpaceX does not want to become a construction company. It wants to put its radios on towers that already exist.

Own the towers, rent the carriers

Here is the part I find most interesting, and it is also the part most people miss.

Carriers come and go. Towers stay.

Think of a landlord and a shop. The shop might change its name, change its owner, go out of business. The landlord still owns the building on the corner, and the next shop still needs a place to open.

Tower companies are landlords for the phone industry. AT&T, Verizon and T-Mobile are tenants. If SpaceX becomes a fourth tenant, the landlord gets a fourth rent cheque. If SpaceX fails, the landlord still has the other three tenants.

That is why Crown Castle and American Tower rose on the same news that hurt the carriers. The market was not saying towers are better than carriers. It was saying the rent just got more valuable, because now there might be one more company fighting to pay it.

I have said this before in a shorter form, and I still think it holds: own the towers, rent the carriers. It is not a slogan. It is just how the money flows in this business.

But let us slow down

I am not saying SpaceX has already won. I am not saying AT&T and Verizon are finished. Anyone who tells you that after two weeks of headlines is selling you a story, not an analysis.

First, nothing is signed. Talks are not leases. Leases are not a network.

Second, the spectrum deal still needs FCC approval. Regulators can slow things down, change conditions, or ask hard questions. That part is not done.

Third, building a carrier is brutally expensive and slow. TD Cowen estimated a nationwide terrestrial network would cost more than $80 billion. That is an estimate, not a SpaceX budget, but it tells you the scale. You need radios, fibre backhaul, power, billing, customer service, stores, and people to fix things when they break. A rocket company can launch satellites. That does not automatically mean it can run a phone shop in Ohio.

Fourth, the carriers will fight back. They already have the customers, the stores, the brands, and the networks. They will cut prices, bundle services, and spend more on their own towers. That is what competition does. It does not just hand the market to the new guy.

And fifth, SpaceX's other spectrum deal, the 2 GHz spectrum from EchoStar, cannot even fully close until around November 2027, according to earlier reporting. So parts of this plan are still more than a year away from being usable.

Bernstein, the research firm, put it well after the first announcement: buying spectrum is not the same as building a nationwide network. I think that is the sentence to keep in your head through all the hype.

So what should we actually watch?

If you want to know whether this is real, do not watch the stock price for a day. Watch for four boring things.

One: FCC approval for the 800 MHz deal. Without that, SpaceX does not own the lane.

Two: a signed lease, not a report about talks. A press release that names the tower company, the number of sites, and the price. That is when renting becomes real.

Three: a real city, with real customers, using a real ground network, not just a satellite test. A demo in a desert does not prove a carrier. A working service in a dense city does.

Four: what the carriers do next. If they start spending a lot more on towers and cutting prices, that tells you they believe the threat, even if SpaceX has not launched anything yet.

Until those things happen, this is a direction, not a destination.

The boring business became the interesting one

I started this essay by telling you to look at a tower on the side of the road. I will end it the same way.

For years, towers were the most boring part of the phone business. Nobody talked about them at dinner. They were just there, holding up antennas, collecting rent quietly while the carriers fought over customers and ads and new phones.

SpaceX changed that, not by inventing a new tower, but by reminding everyone who actually owns the ground the phone industry stands on.

Satellites get the headlines. Rockets get the photos. Spectrum gets the big dollar numbers in the news.

But in the end, your phone still needs something close by, something solid, something that can get through a wall and handle a crowd. That something is usually a tower.

And whoever owns the tower gets paid, no matter whose name is on your phone bill.

That is why the market moved. Not because SpaceX bought 80,000 towers. It did not.

It moved because SpaceX might soon need to rent 80,000 of them. And in this business, the landlord is often the one who wins first.