LESSONS

I learned from my mistake in 1999 that even when you know something is irrational, if the momentum is there, it can keep going long enough to kill you.

I turned $12,415 into $1.65 million. Then I thought I could scale it into a hedge fund. I was wrong. Some edges only work at small size.

When I sold during a market drop in the 1980s, I crystallized a loss and then watched the recovery happen without me. I never made that mistake again.

I lost six billion dollars in one day in 2000. That experience taught me that the conventional rules of business can destroy you overnight. Bitcoin taught me the alternative.

The biggest mistake I made in 2017 was not taking profits. Do not repeat my mistake.

Getting fired from Facebook was the best thing that ever happened to me. I just did not know it for a few years.

We were adding 64,000 rooms a month at our peak. People said it was too fast. They were right. Speed without quality control is just chaos with a growth chart.

We raised $3.5 billion on a vision of the future. The vision was right. The timing was wrong by about a decade. That's an expensive lesson in patience.

If I did it again, I'd start with one city, not fifteen. We scaled before we figured out the math. Classic mistake.

The vision was right. The timing was right. The execution was a disaster. That's the hardest lesson — being right about the future doesn't save you.

I was building a product that the market didn't need as badly as I thought. The checkout problem was real. The willingness to pay for a solution wasn't. That's the lesson.

People will write case studies about us. $120 million in funding. $600K in revenue. It's going to be in business school textbooks under the chapter called "What Not To Do."

The lesson from IRL is simple: if your growth looks too good to be true, audit it. Because your investors won't. They'll just write bigger checks.

I genuinely believed we were building something people wanted. The user numbers looked incredible. I should have asked harder questions about where they were coming from.

Construction is local. Every city has different codes, different unions, different weather, different soil. A factory in Phoenix doesn't help you build in Boston. We learned that the expensive way.

We raised $2 billion and went bankrupt. The idea was right — construction needs disruption. The execution was a masterclass in how not to do it.

I co-founded Leap Motion — a hand-tracking device that was supposed to change computing. It didn't. Midjourney taught me that timing matters more than technology. The tech was ready this time.

Every startup exaggerates a little. But there's a line between optimism and deception. I thought I was being optimistic. The market decided I was being deceptive. The market was probably right.

We shut down in 2020. Laid off everyone. Returned what money we could. The lesson is simple: don't raise on the technology you plan to build. Raise on the technology you've actually built.

Twitter bought us before we even launched. In hindsight, that was both the best and worst thing that happened to Vine. Best because we got distribution. Worst because we lost control.

Vine created an entire generation of creators — Logan Paul, Lele Pons, King Bach. They all left for YouTube and Instagram because we could not pay them. We had the talent and no business model.

TikTok proved the format worked. They just did what we should have done — paid creators, added music, and actually built a business around it.

Sequoia backed us. Temasek backed us. Every name-brand investor in Asia wrote us a check. When that much money comes in that fast, it can make you believe you are infallible. You are not.

Startups glamorize burning cash. Burn rate is celebrated like a badge of honor. It should be treated like what it is — a countdown clock to death.

The mistake was simple — we were spending more on labor than we were charging customers. The unit economics never worked. I just refused to see it until the bank account hit zero.

When Silicon Valley Bank failed on a Friday and we had $3.3 billion stuck there, USDC depegged to 87 cents. Those were the longest 48 hours of my life. The government backstop on Sunday saved us — and taught us to never concentrate reserves again.

I was at Lehman Brothers when it collapsed. I watched a 158-year-old bank disappear overnight. That taught me two things: institutions are fragile, and complacency kills. I run Revolut like it could die tomorrow.

We raised $3.4 billion because instant delivery felt inevitable. The correction taught us that inevitable does not mean immediately profitable.

We were valued at $4 billion. Then people went outside. The lesson: do not confuse pandemic behavior for permanent behavior.

Self-driving is a trust problem disguised as a technology problem. We solved the technology. We failed at the trust.

An 80% stock decline teaches you that growth without profitability is just expensive marketing.

Share buybacks are not a substitute for investing in your business. I learned that at Sears.

Financial engineering cannot substitute for operational excellence. I learned that the hard way.

Being technically better doesn't guarantee winning. Ecosystems win, not specs.

iBuying was our most expensive lesson. Buying houses is not a tech business — it's a capital business.

A brilliant distribution strategy means nothing without product quality. We learned that the hard way.

$14 billion peak valuation to $173 million acquisition. That's the price of ignoring quality.

We proved that people will buy anything if it's cheap enough. That was both our strength and our problem.

Venture-subsidized convenience isn't a business model. It's a temporary subsidy.

We raised $2 billion. In hindsight, that made us feel invincible when we should have felt cautious.

$8 billion SPAC valuation to zero. Cars are not software — every one is a unique snowflake of mechanical risk.

I built Zoopla and LoveFilm — both platforms. Cazoo was a logistics company. That difference killed it.

Every construction tech unicorn that died had the same problem: the technology worked, the industry didn't cooperate.

We pioneered mRNA. Then watched others make $50 billion from it. Being first doesn't mean being best.

Algorithms can't price unique homes accurately enough. Every house is a snowflake of mechanical risk.

Aligning incentives between school and student is easy to pitch and nearly impossible to operate.

The core insight was right: selling a house is painful. The execution was wrong: buying houses at scale is suicide.

$130 million raised, rebranded to BloomTech, then shut down. The idea was right. The execution was wrong.

The panels were beautiful. The logistics were a nightmare. Construction disruption requires more than better products.

Building your business on someone else's platform is playing with fire. We got burned.

Concentration is wonderful when it works. When it doesn't, you learn why diversification exists.

Football is an emotional business. You can't run it like retail. I learned that at Newcastle — painfully, slowly, and expensively.

The connected fitness crash taught everyone the same lesson: hardware alone isn't a business. The software, the data, the coaching — that's the real product.

My biggest regret with Crane Bank was not moving faster to raise more capital when we had the chance. We were growing too fast for our own equity.

Anaconda was my education. It was an expensive education, but I would not trade it. You learn more from failure than from anything else.

The film business is unlike any other I've been in. You can have the right story, the right talent, the right marketing, and the audience still decides. That humility is a good lesson for any investor.

The most important lesson dad taught us was to listen to your customers. Not just what they say — watch what they do.

A network built on token rewards instead of real value is a bubble waiting to pop.

If the game is not fun without the token, it is not a game. It is a Ponzi with graphics.

Your name is the one bankroll you can never rebuy.