Four Questions I Ask Every Week

Most people review their week by looking at what they got done. Tasks completed, meetings attended, emails sent. That measures activity, and you can be busy every single day while going nowhere.

For a while now I've been running a different weekly review. Four questions, about ten minutes, every Sunday evening. No journaling app, no template.

1. Am I becoming more or less free?

This is the question I care about most, and the one most people never ask.

Freedom here means optionality more than money, though money is part of it. Can you say no to things? Can you walk away from a bad situation? Do you have fewer dependencies this week than last week, or more?

Balaji Srinivasan frames this as exit over voice. Instead of trying to fix a broken system by arguing, complaining, or voting, you build the ability to leave. If your job is bad, you don't write a strongly worded email to HR, you build skills and savings until you can exit. If your country's policies are hostile to what you do, you build location independence.

So every week I check whether I added or removed a dependency. A new automated system that runs without me: more free. An obligation that needs my daily attention with no upside: less free. An income stream I don't trade time for: more free. A client who expects me to be available 24/7: less free. The direction matters more than the magnitude. Even a small weekly increase in freedom compounds into a very different life within a few years.

2. Am I building leverage or selling time?

Naval Ravikant's hierarchy of leverage is the most useful framework I've found for this: code, then media, then capital, then labor.

Code scales at zero marginal cost. A trading bot I write once executes thousands of trades without me. Media scales almost as well; a post or a video works while you sleep and compounds as more people find it. Capital scales but you need the base first. Labor doesn't scale at all. If you're selling hours, there's a ceiling, because there are only so many hours.

I run algorithmic trading systems, which makes the distinction concrete. An hour spent improving the system is leveraged, since it trades 24/7 whether I'm awake or not. An hour spent manually watching charts and executing trades is not. The goal is to automate myself out of the loop entirely.

The honest weekly answer is usually a mix. That's fine, but the ratio should be shifting. If you moved from 80% unleveraged to 70%, that's progress.

3. What did I do this week that compounds?

The most honest of the four. Some work accumulates, most of it evaporates.

Compounding work: a trading system that gets better with more data, an audience that grows on its own, a skill that makes every future project easier, a relationship with someone whose trajectory is rising. On the other side: meetings that could have been emails, fixing a bug caused by architecture you never refactored, doing something for the third time that you should have automated after the first. Most of what fills a day gets done and leaves no residue.

The trap is that the non-compounding stuff usually feels more urgent. The inbox screams louder than the long-term project. My bar is deliberately low: one compounding thing per week, which still works out to 52 small assets a year. Some weeks it's a feature for a system that will run for years, some weeks it's a post like this one, some weeks it's a single conversation that reframes how I think about a problem.

4. What's the one thing I'd change?

Not five things, one. If you're allowed to list everything that went wrong, you end up with a vague cloud of dissatisfaction. Forced to pick one, you find the highest-leverage change. Usually it's mundane: I stayed up too late and wasted the next day, I said yes to something I should have declined, I spent three hours avoiding the hard thing. The answer tends to feed back into the first three questions anyway, since the thing I'd change is almost always a moment where I traded freedom for obligation or picked unleveraged work.

Why weekly

Daily reviews are too noisy; one good or bad day means nothing. Monthly is too slow, because by the time you notice a pattern you've already lost four weeks to it. Weekly is long enough to see real patterns and short enough to course-correct.

Fair warning: the answers are sometimes ones you don't want. You'll realize you spent the entire week selling time, or that you're less free than a month ago because you took on commitments without thinking. Uncomfortable, but clarity is the prerequisite for change, and ten minutes on a Sunday is a cheap price for it.