How PayPal was built: lessons from Musk, Thiel, and Levchin

Executive overview

PayPal was created by two companies — Confinity (Thiel and Levchin) and X.com (Musk) — that started as rivals before merging in 2000. For four years, the combined company lurched from one near-death crisis to the next: runaway fraud, hostile regulators, platform risk from eBay, and internal coups.

The founders were mostly in their 20s, deliberately hired for intelligence over experience, and treated speed and impatience as core operating values. What they built beneath the payment product — a real-time fraud detection system driven by machine learning — turned out to be the actual innovation.

The most important lesson: surviving as a startup is not about having the right idea; it is about having the right people, moving fast enough to fix every wrong thing before it kills you.

The founders before PayPal

• Levchin grew up in Soviet Russia with severely limited computer access; writing code with pencil and paper made him tenacious and exacting.
• His grandmother — a physicist who never surrendered — gave him a model for persistence he carried into every crisis at PayPal.
• Musk sold Zip2 for $307 million at 27, then immediately put $12.5 million of his own money into X.com — a recruitment signal that told candidates he genuinely believed in it.
• Musk internalized from physics that framing the right question matters more than finding the answer; he applied this iteratively to every product decision.
• Thiel, meeting investors before Confinity launched, learned that a saturated, undifferentiated market (a smart calendar with 200 competitors) destroys value — the lesson that shaped PayPal's product discipline.
• Thiel and Levchin met when Max wandered into a lecture Peter was giving to five people; Levchin thought: "If I ever do anything in the financial world, that's the person I want to work with."

Recruiting and team building

• Levchin's rule: A players hire A players; B players hire C players. The first B player you hire takes the entire company down.
• All candidates had to meet every member of the team, even at 10 p.m. — one hire was interviewed from arrival until 2 a.m.
• They deliberately avoided hiring people with banking experience; they wanted people who would not say "this is how we did it at the last place."
• To win engineers away from Google ("an oil derrick that spits out $30 billion a year"), Levchin told them: at Google you are a cog; here you are an instrumental piece of something we will build together. Do not compete on salary.
• Early hires included high school dropouts, chess champions, and puzzle enthusiasts — chosen for eccentricities, not despite them.
• Sachs was hired over the team's objections because Thiel valued that Sachs had come in "guns blazing against their flagship product." Candid critics were assets.

The pivot to email payments

• The original product was money transfer via Palm Pilot using infrared beaming. Reid Hoffman tested it by walking into Palo Alto restaurants; the answer was zero to one Palm Pilots per restaurant. The idea was dead.
• The fix was an afterthought: Levchin proposed that the website could send money via email address. When it was first suggested, few saw it as a breakthrough.
• Levchin later admitted he had become an avid user of the afterthought product while remaining committed to the original that he was not using — a warning sign he missed in real time.
• Musk's insight: "We didn't invent money transfer. We just made it useful." Other companies had the idea; none did it right.
• Sachs reduced the signup flow from seven pages to one and put "as easy as email" on the office walls. Every moment of friction was fat to be cut.

The merger and early crisis

• Confinity and X.com independently pivoted to email payments and ended up as neighbors in Palo Alto, launching a weeks-long war to sign up eBay sellers — "a race to see who could run out of money the fastest."
• The merger was opposed by Musk, skeptical of Confinity's lead investor, and almost cancelled by Levchin. Harris, the CEO who forced it through, was quickly ousted.
• Post-merger problems colliding simultaneously: 100,000 unread customer service emails; fraud losses threatening solvency; two incompatible tech stacks; $25 million quarterly burn; ongoing regulatory scrutiny from the FTC and Secret Service.
• Thiel pushed urgency on a $100 million fundraising round because he believed the dot-com bubble was about to burst. Days after the round closed, markets began a slide that wiped out $2.5 trillion in market cap. Five to seven rival payment services died for lack of capital. "If we hadn't raised that $100 million, there would be no PayPal, no SpaceX, no LinkedIn, and no Tesla."

Solving fraud

• At the peak, PayPal was losing tens of millions of dollars per month to fraud — primarily organized rings exploiting stolen credit card details.
• Levchin's team built what became one of the first large-scale real-time machine learning fraud detection systems.
• His reframing: PayPal's actual innovation was not the payment UI but the risk management system underneath — the ability to tell, in milliseconds, whether money being moved belonged to the person moving it.
• Even deliberate losses to fraud were justified: each bad transaction generated data that trained better predictive models.
• PayPal's core business model insight followed from this: if enough customers kept balances inside PayPal rather than withdrawing to a bank, internal transfers cost almost nothing. The debit card and checkbook Musk insisted on were tools to prevent money leaving the ecosystem.

Operating culture and decision-making

• Musk's email to his co-founder during their power struggle: "My mind is always on X by default. Even in my sleep, I am by nature obsessive compulsive. What matters to me is winning — and not in a small way."
• Musk on iteration: "You start off with an idea and that idea is mostly wrong. Then you adapt and keep refining. Too much precision in early plans cuts that iterative loop prematurely."
• Musk and Sachs reorganized into small, self-contained engineering units after reading Frederick Brooks' observation that adding programmers to a late project multiplies communication overhead, not output.
• Cultural operating principle: "Be intolerant of slowness." Company leaders set a tone of impatience and made decisions by fiat when necessary.
• On the value of young, inexperienced teams — Musk on the "seasoned executives" from banks: "These are the same executives who can't compete with us. Doesn't make sense."

Platform risk and the eBay fight

• eBay launched "Buy It Now" with its own payment service (Billpoint) as the default, cutting PayPal out of the checkout flow.
• Thiel tasked Keith Rabois with building an antitrust paper trail against eBay, citing the Microsoft case; PayPal also created a political action committee and enlisted outside legal counsel to write formal complaints.
• Billpoint gained only marginal share because PayPal was solely focused on payments while eBay was managing an entire marketplace; focused beats distracted in a product fight.
• Guerrilla marketing at eBay Live: PayPal distributed thousands of T-shirts to sellers with a cash prize for wearing them during Meg Whitman's keynote. Whitman walked onto stage facing a sea of PayPal logos.
• To silence eBay during the IPO quiet period, Hoffman re-opened fake acquisition talks at $1 billion, knowing eBay could not publicly criticize a company it was in negotiations to buy. "My principal goal was to keep them quiet and not to sell the company."

The coup against Musk

• A clandestine group planned Musk's ouster while he was on a joint honeymoon and fundraising trip; they timed the board meeting for when his plane took off.
• Musk's response was to instruct allies not to resign in protest: "While I didn't agree with their conclusion, I understood why they took the action they did. Peter and Max and David are smart people with generally good motivations."
• His reasoning: "I view the company at least partly as my baby. If I attack the company and the people there, it's like I would be attacking my baby."
• Just months after being ousted, Musk told a friend over drinks: "I'm going to colonize Mars. My mission in life is to make mankind a multi-planetary civilization." He immediately put his PayPal payout into SpaceX and Tesla.

The IPO and what came after

• PayPal IPO'd in February 2002, during a period when the press uniformly advised investors not to buy. eBay acquired the company for $1.5 billion later that year.
• Musk, as the largest individual shareholder, received approximately $180 million — which he immediately deployed into SpaceX ($100M) and Tesla ($70M).
• Looking back, the founders identified their inexperience as an asset: "Had the company built its fraud process traditionally, they would have hired people who had been building logistic regression models for banks for 20 years but never innovated."
• Levchin's observation on the best employees: "Many of its earliest employees simply hated being employees. The very best employee at any level is the person who believes this is their last job working for someone."
• Reid Hoffman's lasting habit from the PayPal years: he regularly asks contacts, "Who is the most eccentric or unorthodox person you know and how could I meet them?" He is searching for founders who resemble his once less-than-perfectly-polished colleagues.
• Levchin on what defined that group: "One key ingredient is an almost irrational lack of fear of failure and irrational optimism. They managed to not get caught up in all the little details while being remarkably aware of the really important ones."