The Four Leverages

This is the reading version of the Four Leverages module. The module is fast and interactive. This is the slow, deep version, for when you want to sit with the idea properly.

Two accountants, same skill, different lives

Two accountants. Same training, same city, same starting point. One files returns by the hour. She is good, her clients trust her, and after twenty years she is still trading each hour for one fee. The day she stops working, the income stops with her. The other built a small piece of tax software that quietly files returns for ten thousand people while she sleeps.

Nothing about talent separates them. Nothing about effort either; if anything, the hourly one works harder. The entire gap is one structural choice: whether your output is chained to your hours or set free from them. Skill decides what you can do once. Leverage decides how many times it pays you.

That word, leverage, gets thrown around loosely, so let me pin it down. The framework here is Naval Ravikant's, his map of the four kinds of leverage, and everything below is his idea in my own words. I say that upfront because the map is too good to quietly steal.

Leverage is a multiplier on your judgment

Strip the idea down to its mechanic. When you make a good decision, leverage is whatever multiplies the result. A good call with no leverage helps one client. The same call, wired into a product or a piece of content, helps a hundred thousand people and costs you nothing extra to repeat.

Most people spend an entire career sharpening their judgment and never once ask what that judgment is plugged into. Each right call keeps helping exactly one person, one project, one boss. A great decision with no leverage is a great decision that helped one person. That is the quiet tragedy under most careers, invisible because the judgment really is improving. The multiplier just stays stuck at one.

Naval's map says there are four things your judgment can plug into. Two are old. Two are new. The difference between those pairs might be the most important economic fact of your working life.

The old levers: labor and capital

For most of history, if you wanted to multiply your output beyond your own two hands, you had exactly two options. You could get people to work for you, or you could get money to work for you. A factory owner used both at once: workers on the floor, capital in the machines.

Be fair to these levers, because dismissing them is fashionable and wrong. A team that runs without you is leverage most founders would kill for. Money compounding quietly in good assets has made more fortunes than any app ever will. Labor and capital still build empires.

But notice what both of them demand before they move an inch. It was never money itself; money could be borrowed from a bank or charmed out of an investor. It was never skill either; plenty of brilliant foremen and clerks never owned a factory, because nobody handed them the workers or the machines. The thing you could not borrow your way around was the yes. Labor means people consent to be managed by you. Capital means someone writes you a cheque or a loan. Both levers only move after a person who already holds power lets you in.

That is the toll booth. To pull the labor lever you take on payroll, management, and the weight of carrying people. To pull the capital lever you need money you do not have, or you must persuade someone who does. For most of a working life, both doors stay locked, and they stay locked hardest for the people who need leverage most: the young, the broke, the unconnected.

The new levers: code and media

Now consider a writer who spends one weekend recording a course on negotiation. She uploads it once. Over the next three years, forty thousand people buy it. She never reopens the file. Compare her to a consultant who is genuinely brilliant at negotiation and bills by the day.

Notice what she never needed. No boss to hire under. No investor to fund her. Not a single employee. One person made the thing alone, on a laptop, and the world could use it that same night. The accountant's tax software from the top of this essay has the same property. These are the two new levers: code, meaning software and products that run while you sleep, and media, meaning content that copies itself to anyone at zero cost.

The historic thing about them is not that they are modern. It is that they are permissionless. There is no interview, no pitch meeting, no approval to win. Cheapness alone is not the revolution either; cheap but gated things have always existed, and a radio slot costs little to run yet still needs a licence. What is new is that nobody stands at the gate at all. A teenager and a billionaire reach for this lever on identical terms, which no older leverage ever allowed. You do not wait for anyone above you to say yes. You grant the permission to yourself.

Zero marginal cost is the whole secret

Permissionless is half of it. The other half is a boring accounting term that quietly runs the modern economy: marginal cost, the cost of serving one more person.

A barber has high marginal cost. One more haircut is one more head, one more half hour of his life. Our brilliant consultant has the same problem in a nicer shirt: his tenth client costs him a tenth day. Every buyer costs him a slice of his life, so his income is capped by his calendar.

The course creator's ten-thousandth buyer costs her nothing. A song on Spotify has the same property: the ten-millionth stream costs the artist exactly what the first one did: nothing. When marginal cost falls to zero, your ceiling comes off. You stop asking how many hours are in a day and start asking how many people are in the market. Every durable fortune of the last twenty years sits on top of something that could be copied for free.

People sometimes credit reach instead: the internet lets you sell worldwide, so reach must be the advantage. It is not; it sits downstream of the real one. The consultant could market himself to the whole planet tomorrow, and every client he won would still cost him a day of his life. The cap was never how many buyers he could find. It was what each one cost him to serve.

Why one person can now beat a company

Put permissionless and zero marginal cost together and you get the strangest economic fact of our era: a single person with code or an audience can out-leverage an entire company.

Look at what the individual skips. No payroll, so no drag from managing and carrying people. No investor, so no cap table and no asking. No per-unit cost, so reaching a million people takes the same effort as reaching ten. The company hauls all three weights and competes anyway. This is not a story about the loner outworking the firm. It is about the loner pulling a lever that the firm's own structure makes clumsy. An army of one, if the one has code or an audience. That line is mine, but I am standing squarely in Naval's lane when I say it.

The objections, taken seriously

Every seductive framework deserves its cross-examination, so let me argue against myself for a minute.

First objection: survivorship bias. For every course that sells forty thousand copies, thousands sell nothing. True, and nothing in this essay promises you a hit. But under the old levers, even the attempt needed a yes from someone with power. Under the new ones, the attempt itself is free of gatekeepers, the downside is a few evenings or a few lean months, and the upside is uncapped. You are not guaranteed a win. You are guaranteed a turn, and for most of history you were not even guaranteed that.

Second objection: hourly work pays this month, products pay maybe, someday. True again, which is why the lesson is not to fire your clients tomorrow. Steady cash is oxygen; keep breathing. The lesson is narrower: hourly work, however well paid, has no multiplier, so some fraction of your time has to be spent converting skill into something that pays more than once. Keep the shift. Build the asset on the side of the shift.

Third objection: if anyone can pull these levers, will they not get crowded until nothing is left? Access did get crowded. Judgment did not. Leverage multiplies bad decisions with the same enthusiasm as good ones, which is why most attempts fail on judgment, not on access. A world where everyone holds a lever is a world where good judgment is worth more than ever, not less. The gate did not move to the front of the market. It moved inside your head.

Fourth objection: the biggest empires still run on labor and capital. Correct, and the framework agrees. The answer is sequence, not rejection; we will get there.

What the two developers teach

Here is the sharpest version of the choice. A skilled developer takes a contract at forty thousand a month. He is good, he is busy, he is fully booked. A friend with the same skill spends three months building a small paid tool instead, earns nothing at first, then settles into thirty thousand a month that arrives whether he codes that day or not. A year in, who is stronger?

The seductive answer says the contractor: he can bank the extra ten thousand a month and buy compounding assets with it. That works only if the surplus actually becomes assets, and a fully booked contractor is exactly the person who never makes that switch. Meanwhile his entire income still dies the day he stops or the client walks. It is a bigger number with no machine under it.

The cautious answer says it is too early to call, since one year of product revenue proves less than a contract that keeps renewing. Fair caution, wrong lens. If the tool dies, the builder still owns the skill plus a proven playbook for the next one. If the contract dies, the contractor owns nothing but his next billable hour. One of them compounds even through failure, and that asymmetry is the whole answer.

The second-order effects are starker still. Once income detaches from hours, time itself becomes investable. The builder can stack a second bet on top of the first, raise prices, or simply rest, and the machine keeps paying. Each product also makes the next one cheaper to build, because the audience, the playbook, and the trust carry over. Leverage does not just multiply income. It compounds sideways into every future attempt.

How to actually choose your lever

A working order, not a vibe.

If you have no money and no team, start with the permissionless two, code or media, because you can begin tonight without asking anyone. Then pick by temperament. If you build, ship a product. If you explain, build an audience. Do not agonise over this fork; you are choosing a starting lever, not a life sentence.

Media often comes first even for builders, because it is the cheapest lever to start and it creates the demand your code or product later fills. An audience that trusts you is a queue standing outside a shop you have not opened yet.

Then, once cash flows, layer the old levers back in deliberately. Hire labor to remove yourself from the work you have already systemised. Deploy capital so money compounds beside you while you build. The sequence runs permissionless first, permissioned second: earn the leverage nobody can deny you, then buy the leverage that once required a yes.

And here is the exercise, this week, not someday. Look at how you earn today and find the lever under it. If your income stops the day you stop working, you are on labor with no multiplier, and that is most people, so no shame, just clarity. Now pick one permissionless move you could start this week without anyone's approval. If you build, name one small product that could turn your skill into something that runs without you. If you explain, name one piece of media that captures what you know once and reaches people while you sleep. Then write the single sentence: this week I will build or publish ___, so that I earn from it more than once.

If your income dies the day you stop, you do not have a business. You have a shift. The lever does not pull itself. But for the first time in history, nobody can stop you from reaching for it either.

Want the fast, interactive version instead? Run the Four Leverages module, or explore the whole codex.