SpaceX landed Starship for the first time and I

SN15 landed.

I need you to understand how much those two words mean.

SpaceX has been flying Starship prototypes from Boca Chica, Texas, and every single one has exploded. SN8 flew beautifully, did the belly flop, did the flip, and hit the ground too hard. Fireball. SN9 did basically the same thing. Fireball. SN10 actually landed, stood upright for about ten seconds, and then exploded. SN11 blew up in mid-air during the flip, in fog, and rained debris on the landing pad.

I watched all four of them. I wrote about SN8. Each time, the explosion was expected, even celebrated (the data was good! the belly flop worked!). But each time, there was also a small part of me that wanted to see it land. Just once. Just to know it could.

Today it did.

SN15 launched at 5:24 PM CDT. It climbed on three Raptor engines. The engines shut down in sequence. The ship tipped horizontal. Fell. The aerodynamic flaps guided it down. At 500 meters, the engines relit. The ship flipped from horizontal to vertical. And it landed.

Gently. Upright. On the pad. In one piece.

I yelled. I mean I physically yelled, alone in my apartment, at a NASASpaceflight livestream on my laptop. I yelled loud enough that my neighbor knocked on the wall.

I paused the stream, went to her door, and showed her the replay on my phone.

“It’s a rocket,” she said.

“It landed,” I said.

“OK,” she said.

She didn’t get it. That’s fine. I get it. After months of fireballs, that gentle touchdown on the pad was one of the most satisfying things I’ve ever seen from a screen.

Why this matters

Starship is designed to be fully reusable. The whole ship. Upper stage and booster. No part gets thrown away. That’s never been done before. The Space Shuttle was partially reusable (the orbiter came back, the boosters were recovered, the external tank was discarded). Falcon 9 lands its first stage. But a fully reusable orbital-class rocket? Nobody has done that.

If Starship works, the cost of getting to orbit drops dramatically. SpaceX is targeting around $10 per kilogram to low Earth orbit. For context, the Space Shuttle cost roughly $54,000 per kilogram. Even Falcon 9 costs about $2,700 per kilogram.

$10. Per kilogram.

At that price, space isn’t just for NASA and billionaires. At that price, you can put a factory in orbit. A hospital. A hotel. A school. The economics change so completely that the question shifts from “what can we afford to put in space?” to “what shouldn’t we put in space?”

The iteration speed

Here’s what continues to amaze me.

SN8 flew in December. SN15 landed in May. Five months. Four prototypes. Four explosions. One landing. That’s a development pace that the aerospace industry has never seen.

Traditional aerospace builds one vehicle, tests it for years, certifies it, and then flies it. SpaceX builds multiple vehicles simultaneously, flies them, learns from the failures, and flies again. In weeks.

The result is visible: each prototype was better than the last. SN8 got the belly flop right. SN9 was basically a repeat for data collection. SN10 almost landed. SN11 had an engine issue during the flip. SN15 got everything right.

Five months from fireball to landing. In rockets. Where the traditional timeline for that kind of progress is measured in decades.

The small fire

There was a small fire at the base of SN15 after it landed. A methane leak, probably. Ground crews sprayed it with water and it went out. The ship stood there, slightly scorched, slightly leaning, absolutely intact.

That image, a Starship standing upright on the pad with a small fire licking at its base, feels like a metaphor for the whole program. Not perfect. Not clean. A little on fire. But standing. And ready to fly again.

I watched the replay seven times.

My neighbor hasn’t knocked again.

Related thinking: