The four Cs of leverage, and how to use them when you have none

In November 2023 I got twenty minutes with Sahil Bloom, whose writing I had followed for years.

I spent some of it telling him about a problem I was embarrassed by. I had done things at fifteen that I could not seem to repeat, and I felt permanent pressure to match them. His answer reframed it: rather than treating what I had already done as a standard to live up to, treat it as something to use.

It took me a few weeks to see what that meant. I had been running from my own history, trying to become someone else, at that point attempting to start a tech company I had no particular reason to start. What I actually needed was to point what I already had at the thing I actually wanted.

Which raised a more practical question. What does using it actually mean?

That took me to leverage, and to a framework I first found in the Almanack of Naval Ravikant and then in a version from Alex Hormozi. They both write about it in a business context. It applies more widely than that.

Start with the word

Leverage comes from lever. A lever is a tool that lets you move something heavy by applying force somewhere else. The Egyptians used them to lift stones weighing over a hundred tonnes.

The useful property is disproportion. The output is larger than the input. So the question for anything you are building is simple: where are the levers, and are you using any of them, or are you just pushing on the rock.

There are four.

First C: collaboration, other people's time

When you start something you trade your own time for outcomes. Money, impact, grades, whatever you are after. The ceiling arrives quickly, because there are twenty four hours in a day and some of them have to be spent asleep.

Other people are the first lever past that ceiling.

At the nonprofit I started, it took us three years to reach five hundred people. Then the pandemic made our original approach impossible, and we were forced into working with organisations already operating across Kenya who could distribute what we made.

Five hundred to eight thousand in a single year. Same product, same idea, different mechanism.

Whether it is a school project, a job, or something you are building on your own, the question is who else could carry part of this.

Second C: capital, other people's money

Money makes everything faster, and you usually do not have much of your own.

When I started at fifteen I asked a lot of people to donate and collected a lot of rejections, for a fairly reasonable reason: very few people want to hand money to a fifteen year old with no track record.

So we sold shawarmas at school events. Two or three hundred dollars an event, every couple of weeks. That worked, and it also had an obvious ceiling, because there are only so many events and only so many shawarmas.

The change came from asking for something other than charity. We started pitching companies on sponsoring our products in exchange for their logo appearing on them. That is a trade, not a favour, and people respond to it completely differently. The first one was twelve and a half thousand dollars. The second was thirty thousand.

That money is what took us from five hundred people to twenty five thousand.

The permission line

Naval makes a distinction here that is worth stopping on.

Collaboration and capital both require permission. Someone has to decide to work with you. Someone has to decide to give you money. You can improve your odds, but you do not control the outcome, and when you are young and unproven the odds are poor.

The next two require nobody's permission at all. That is why they matter most when you are starting.

Third C: content, other people's attention

You can reach an enormous number of people at close to zero cost, which is genuinely new. When I first wrote this I was twenty two, writing to more than a hundred thousand people from a dorm room in Vancouver, and the marginal cost of one more reader was nothing.

Here is the clearer example. In 2020 I joined Sophia Kianni as an outreach director at Climate Cardinals, which she had just founded. The first problem was recruiting volunteers.

TikTok was at its peak, so she asked friends to make videos about the application. They went viral, and we received over six thousand applications.

That volume then became leverage of a different kind. It was the basis for partnerships with organisations including National Geographic, Translators without Borders and UNEP, because six thousand volunteers is a fact that changes the conversation.

Content does not require you to think of yourself as a creator. It requires something that keeps working when you are not in the room.

Fourth C: code, a machine's time

Code scales in a way nothing else does. You build something once and it runs while you sleep, and each additional use costs approximately nothing.

Most of the largest companies of the last few decades were built on exactly this property. With current tools, the amount of it available to someone with no engineering background is far greater than it was even a few years ago.

The question is which repeated thing in what you are doing could stop requiring a person.

Use it as a diagnostic

You do not need all four, and trying to run all four at once is how people end up doing none of them properly.

When I wrote this, Leaders of Today was using no outside capital at all. It was collaboration, content, and some code underneath. That was a deliberate choice about which constraint mattered most at that point, not a gap.

So the useful version of this framework is a question rather than a checklist. What is actually limiting the thing you are building right now? If it is your own hours, that is collaboration. If it is money, that is capital, and the trade version usually works better than the ask. If it is that nobody knows you exist, that is content, and you can start this week without anyone's approval.

If the honest answer is that you have not started, leverage is the wrong problem to be solving, and how to actually take the first step is the right one.

Pick the one that matches your actual constraint. Ignore the other three until it changes.