# SpaceX's initial network could cost $10–15 billion

Markdown copy of https://realalpha.app/s/spacexs-initial-network-could-cost-10-15-billion/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy.

- Speaker: [Walter Piecyk](https://realalpha.app/people/walter-piecyk/) · Walter Piecyk is a telecom analyst at LightShed Partners. · Covers wireless networks, spectrum and smartphones as a LightShed Partners partner and TMT analyst.
- Host: [Nilay Patel](https://realalpha.app/people/nilay-patel/) · Editor-in-chief, The Verge · Helped start The Verge in 2011 and has hosted Decoder since 2020.
- Show: Decoder
- Episode: [Why Elon Musk wants to sell you cell service | Decoder](https://realalpha.app/v/a9jnh4Vb-GY/), Oct 17, 2026
- Moment: 13:24 to 18:03 of the episode (5 min), playing on https://realalpha.app/s/spacexs-initial-network-could-cost-10-15-billion/; the episode from there: https://realalpha.app/v/a9jnh4Vb-GY/?t=804

## Quote

> “You're probably talking like $10 billion, $15 billion for **that initial build** in terms of CapEx.”
> — [Walter Piecyk](https://realalpha.app/people/walter-piecyk/) at [17:41](https://realalpha.app/v/a9jnh4Vb-GY/?t=1061) · Decoder

## Takeaways

1. Satellite coverage can help differentiate service in underserved areas.
2. A wholesale agreement could help SpaceX enter sooner.
3. Piecyk estimates $10–15 billion for the initial build.

## About this segment

Walter Piecyk argues SpaceX can enter with a smaller build than the incumbents, using satellite coverage alongside terrestrial service. He estimates the initial network could cost $10–15 billion.

From “Why Elon Musk wants to sell you cell service | Decoder”, Oct 17, 2026. [Watch the full episode](https://realalpha.app/v/a9jnh4Vb-GY/)

## Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

[13:34](https://realalpha.app/v/a9jnh4Vb-GY/?t=814) So they're going to overcharge them for that. I'm just stuck at, that's a lot of cost, even if you're Elon Musk and you can make some set of people buy whatever you're selling. The third thing you said is obviously the most challenging. That's the thing, that's the kind of the tent in the pole that prevents you from launching services sooner rather than later. But I think there are, without necessarily getting a wholesale agreement, there might be other ways roaming agreements exist out there.

[13:58](https://realalpha.app/v/a9jnh4Vb-GY/?t=838) There are roaming partners that exist that these telcos still rely on. It's very expensive. It's not something you can build a high margin business on, but it perhaps provides an entry to the market sooner rather than later. And then the point of differentiation ultimately becomes, you know, not getting as Apple has pitched in terms of satellite connectivity, some person in the middle of the Utah desert texting because, you know, they're injured.

[14:23](https://realalpha.app/v/a9jnh4Vb-GY/?t=863) There's value there, but it's the soccer field. Like we have, you know, maybe not in the cities, but in the suburbs, there are plenty of areas that the telcos have effectively given up on trying to provide coverage because of NIMBYism, meaning people like just don't want these cell sites, that this satellite coverage cannot just give you this basic text messaging that Apple has been doing through the Global Star constellation, but a true cell phone type experience.

[14:49](https://realalpha.app/v/a9jnh4Vb-GY/?t=889) So A, that's a positive. And the second differentiation is the fact that these telcos have 60% EBITDA margins. So there's a lot of fat there that can be eaten off of that when you're selling other services, you know, in terms of Elon, like selling home broadband, maybe this becomes. It was an entree in terms of pushing Grok and people's usage. You don't necessarily need 60% margins, you know, in this case for him, you know, to justify building this network.

[15:19](https://realalpha.app/v/a9jnh4Vb-GY/?t=919) They also talk about, you know, other data connectivity that these, you know, satellites are going to enable. Again, the current constellation, you need a lot of power and you need a dish. But if you don't have that, if you have the cell phone type power, then the kind of IoT and other type of opportunities that exist there become, I think, another thing that they look at as a revenue opportunity.

[15:40](https://realalpha.app/v/a9jnh4Vb-GY/?t=940) It's just funny, the IoT story, I mean, that's what all these companies promised us about 4G, that's what they all promised us about 5G. Like, there's something about this business that just rhymes every time. I'm confident we're gonna hear about self-driving cars and robot surgery when it comes to Starlink and Tesla and Optimus. Like, there's something about these applications for next generation networks that just recur and we hear about them all the time.

[16:02](https://realalpha.app/v/a9jnh4Vb-GY/?t=962) And then I think about our consumer audience at The Verge, and they want long battery life and fast service and all the places they go. And I'm not sure how that squares with your phone is going to connect directly to a satellite in the middle of a soccer field, right? Like there's something about that that also seems challenging. I think it's just the performance in that if your phone does connect to the satellite on the soccer field, and it works, right?

[16:24](https://realalpha.app/v/a9jnh4Vb-GY/?t=984) I mean, again, a lot of the things that people said wouldn't work on Starlink in terms of airplanes or even getting CPE down below a certain price, they've delivered on. Obviously, Elon sometimes delivers on them a little later than some of his initial optimistic promises, but they ultimately get delivered on. I think there is the ability with these new constellations to give you that type of connectivity on these spaces.

[16:50](https://realalpha.app/v/a9jnh4Vb-GY/?t=1010) Is that enough to differentiate with consumers? I don't know. The telcos have attacked this and saying, well, no one cares about our D-to-D service. Well, your D-to-D service is not good, right? It's just sending basic text messages. What they're talking about is giving you that blanket coverage where you get what you expect from your cell phone wherever you go. That's fine, but you still have to give it to the person in the middle of the building where their cell phone works today and obviously provide the volume at a cost that is going to be attracted to that consumer.

[17:21](https://realalpha.app/v/a9jnh4Vb-GY/?t=1041) So those are, again, longer-term things that are going to require some amount of terrestrial network build. Then the question is how much, right? And I think a lot of the attacks that have been out there have been, you know, referencing $100 billion, $200 billion. I get that. Why? Telcos, they spend, each of them, $10, $15 billion. a year. But they are maintaining massive customer bases. To get that initial network up where you have that usage, where you use that low-band spectrum to get in the buildings, is not those types of numbers. You're probably talking like $10 billion, $15 billion for that initial build in terms of CapEx.