Leverage: The Mental Model of Labor, Capital, Code and Media

What is leverage?

Leverage is anything that lets a small input produce a large output. As a mental model, it means looking at any piece of work and asking how much it multiplies you, and what it would multiply if you were wrong.

The image is physical. A lever lets a person lift a rock far heavier than they could move by hand. The person did not get stronger. The arrangement changed: a rigid bar, a fixed point to pivot on, and force applied far from that point.

In work and life, the bar and the fulcrum are other things:

• Labor: people who act on your decisions.
• Capital: money that buys assets, time or other people’s work.
• Code: software that does a task again and again without you.
• Media: writing, audio and video that reach people while you sleep.

The first two need permission: someone has to hire you, lend to you or invest in you. The last two mostly don’t. That difference shapes who can use which lever.

Where it comes from

The lever is one of the simple machines described by Greek mathematicians. Archimedes (3rd century BC) worked out the law of the lever: force times distance on one side balances force times distance on the other. The later writer Pappus of Alexandria credits him with a famous line, usually rendered as “Give me a place to stand and I will move the earth.” Exact wordings differ across sources.

Finance borrowed the word for borrowed money: a company or investor using debt is “leveraged”, because gains and losses on the full position land on a smaller slice of their own money.

The four-way split of labor, capital, code and media comes from the investor Naval Ravikant, who laid it out in a widely shared 2018 thread on building wealth. His point was that code and media are “permissionless” leverage and that they have become the main way individuals scale their judgment.

How to use leverage

1. Name your current multiplier. For each main activity, ask: if I stopped, how long would the output keep coming? Hourly work stops instantly. A tool, a book or a trained team keeps going.
2. Find the fulcrum first. Leverage on the wrong point wastes force. Use the Pareto principle to find the few inputs driving most results, and the theory of constraints to find the bottleneck that limits everything else. That is where a lever pays.
3. Start with permissionless leverage. Write the guide once instead of answering the same question fifty times. Automate the repeated step. Record the explanation. These need only your time.
4. Add labor and capital when judgment is proven. Hiring and borrowing multiply whatever process you have. Make sure it works at small scale first.
5. Size leverage to survive being wrong. Ask what happens if your central assumption fails. Keep a margin of safety big enough that a bad outcome hurts but doesn’t end the game.
6. Improve your judgment, because that is what gets multiplied. Leverage turns the quality of your decisions into the size of your results, in both directions.

A simple way to compare options is to estimate the ratio of output to your own hours for each. Answering one customer’s question has a ratio of one. Writing an answer that a thousand customers read has a ratio far higher, for the same hour. You won’t get exact numbers, and you don’t need them. The rough ranking is usually obvious once you write it down, and it shows which work deserves your scarce, focused time.

Also note which levers you control. A lever you rent (an ad platform, a single big client, a marketplace) can be taken away. A lever you own (your list, your code, your skills) stays with you.

3 real-world examples

Business

Say a consultant answers the same onboarding questions in every client call. She turns the answers into a short written guide and a recorded walkthrough. Now each new client gets the same quality without her time, and her calls go to the hard problems only she can solve. Nothing about her skill changed. She moved it from labor (her hours) to media (the guide), which can be used a thousand times at no extra cost.

Personal life

Imagine someone who spends Sunday evenings planning meals, then shops once and cooks in batches. One hour of planning saves many hours of daily decisions and last-minute takeout. It is small-scale leverage: one well-placed input controls a week of outputs. The same logic applies to good routines, checklists and defaults.

Tech

A developer writes a script to clean a messy data export that used to take a colleague two hours each week. The script took a day to write and now runs in seconds. Over a year, that day returns roughly 100 hours. Code is the purest leverage most people can access: once written, it runs at near-zero marginal cost. The same applies in reverse: one bad line deployed widely can break things for everyone at once.

Common mistakes

• Leveraging an unproven process. Hiring or spending to scale something that doesn’t work yet only scales the losses.
• Ignoring the downside multiplier. Borrowed money, a viral post and a large team all magnify errors. Many investors were ruined by debt, not by bad picks.
• Confusing busyness with leverage. Managing more people or more tasks is not leverage if the output per unit of your judgment doesn’t rise.
• Waiting for permission. Many people wait for capital or a team while code and media sit unused.
• Pulling the lever in the wrong place. Effort applied to a non-bottleneck is wasted, however clever the tool.
• Renting all your leverage. Building only on a platform you don’t control means one policy change can remove the lever overnight.

Leverage vs compounding

The Mentarium take

Leverage is the model behind most outsized results, and it is also behind most spectacular failures. The difference is rarely the lever. It is the judgment applied to it. So we treat leverage as a two-question check: “how much does this multiply me?” and “what does it multiply if I’m wrong?”

The bias it combats is overconfidence. People take on debt, hire fast or broadcast a half-baked idea because they are sure it will work. Leverage makes that certainty expensive when it is misplaced. Before adding leverage, run The Mentarium Decision Loop (Frame → Map → Stress-test → Decide → Review), and give the Stress-test step real weight: what if demand is half your estimate, or the key hire fails?

Our practical advice: begin with code and media. They are cheap, permissionless and forgiving. Build a small body of work, see what people actually use, then add labor and capital to what has earned it. That order keeps the downside small while your judgment improves, which is the only thing leverage can’t supply for you.