How Elon Musk built SpaceX from nothing to orbit

Executive overview

SpaceX started with $100M of Musk's own money, an empty factory, and a handful of engineers. The goal was simple: lower the cost of reaching orbit, because without that, Mars was impossible.

Three rockets failed before the fourth reached orbit. The company nearly died multiple times. What kept it alive was a combination of extreme speed, brutal hiring standards, and an owner who refused to accept that failure meant stopping.

The core insight: companies are effective problem-solving machines — and SpaceX survived because Musk treated every obstacle, financial, technical, or bureaucratic, as a solvable problem rather than a reason to quit.

Hiring and talent

• Musk personally interviewed every hire through the first 3,000 employees, including weekends and late nights
• His filter: brilliant, hard-working, no nonsense — "there are a ton of phonies out there"
• He prioritised young, hungry engineers over legacy aerospace experience
• Early hires received large equity stakes to align incentives with frugality
• SpaceX attracted missionaries: half of one top graduate program's best students took pay cuts to join before the company had proven anything

Speed as strategy

• Legacy aerospace moved at committee speed; SpaceX operated at founder speed
• Offer letters sent at 1am; a new VP of machining was asked to start the same Saturday evening he was hired
• Musk collapsed engineering and spending decisions into one head: "My brain trusts itself"
• No plans required — Shotwell was told "don't talk about doing things, just do them"
• Iterative design over linear: build, test, find failures, adapt — rather than years of qualification before touching hardware

Vertical integration and cost

• When a machinist supplier threatened to go out of business, Musk bought their machines and brought the team in-house on a Saturday afternoon
• In-house machining cut manufacturing costs roughly in half
• Engineers could talk directly to machinists in the same room instead of routing through buyers and callbacks
• The incentive structure meant employees treated the company's money as their own

Fighting the establishment

• The Air Force initially placated SpaceX assuming it would fail like every previous private rocket startup
• When the Air Force blocked their California launch site indefinitely, Musk moved the entire operation to a remote Pacific island within days
• SpaceX sued Northrop Grumman, Boeing, and Lockheed Martin within its first three years
• Protested a NASA contract others advised against challenging; the Government Accountability Office sided with SpaceX
• Musk's view: "If we don't fight this, we're doomed anyway"

The four launches

• Launches 1, 2, and 3 all failed; the company had parts for exactly one more rocket
• After Flight 3, Musk had no house, was divorcing, and both SpaceX and Tesla were nearly out of cash
• Rather than assign blame, he gave a brief speech: "Get your shit together, go back to the island, and launch it in six weeks"
• The team rebuilt the rocket on the island in days after it crumpled mid-flight on the C-17 transport
• Flight 4 reached orbit; the celebration was matched only by the relief
• A NASA contract worth $278M arrived shortly after, buying the company enough runway to survive

Musk's mental models

• Read biographies of previous rocketeers to learn from their failures before starting
• Studied Andy Beale, who put in $200M and failed — and decided the lesson was about bureaucracy, not technical limits
• After each failure, placed it in historical context: early Atlas had 9 successes in 20 attempts; Soyuz 9 in 21
• Introduced "always go to 11" after a known but deprioritised risk (fuel slosh) destroyed Flight 2
• On reusable rockets: "It's stupidly difficult. It would be one of the biggest breakthroughs in the history of humanity. That's why it's hard."