I Am More Interested in Leverage Than Productivity

I had a genuinely productive Thursday last month. I remember it mostly because of how little it changed.

Everything got answered. Nothing was left open. I closed the laptop with that clean, slightly smug feeling you get when the list is empty.

Then Friday arrived and looked exactly like the Thursday before it. Same shape of work, same volume, same me. I had been fast. I had not been different.

That gap is the thing I keep circling back to. Productivity asks how much I can get done inside an hour. Leverage asks what that hour is still worth after it ends.

One note on the word before I use it forty more times. Take it in the physical sense — a lever and somewhere to stand, so the same push moves more weight. Not a metaphor about money. A metaphor about force.

Let’s imagine two people with the same job and the same Tuesday.

1. The first one is excellent. She clears the entire queue by six — every message answered, every call logged, every commitment honored. Not one thing dropped.

2. The second one clears most of it, then spends the last hour of the day writing down how she decided which things mattered and which could wait. Her queue ends the day slightly untidy.

On Tuesday evening, the first person is clearly ahead. She got more done. If you measured them, you’d measure her as the better operator, and on that day you’d be right.

Now run the same week forward six months. The first person’s Tuesday is finished — it produced exactly one Tuesday’s worth of output and then it was over. The second person’s last hour is still working. It became a rule she doesn’t re-derive, a thing she can hand to someone else, a decision she stopped making from scratch every morning.

That’s the whole distinction. Speed doesn’t accumulate. Structure does.

The treadmill nobody admits is a treadmill

Here’s the part that took me an embarrassingly long time to notice: productivity has no finish line built into it.

The reward for clearing your list is a longer list. The reward for getting faster is a higher expectation of your speed. Whatever you can now do in an hour becomes the new baseline against which the next hour is judged — by other people, but mostly by you.

I don’t think that’s a character flaw or a burnout story. It’s just the shape of the thing. Throughput is measured per unit of time, so it resets every unit of time. You wake up at zero. You always wake up at zero.

And the specific tiredness that comes from it isn’t about volume. It’s about the fact that none of it compounds. A brilliant week and an average week both end on Friday, and by the following Wednesday you cannot tell which one you had.

Picture someone crossing a field of tall grass. They can train, get stronger, learn to move through it faster — that’s real improvement, and it’s gone the moment they stop. Or they can walk the same line enough times, deliberately, that a path appears. After that, everyone crosses the field faster, including the version of them who’s exhausted and having a bad morning.

Productivity is moving faster through the grass. Leverage is the path.

What I built, and what it wasn’t

The clearest example I have from my own work is the agent system that runs my sales pipeline. It’s the thing people ask me about most, and it’s the thing I most often describe wrongly when I’m not careful.

The wrong description: I got faster at the work.

I didn’t. I’m not quicker at writing follow-ups than I was two years ago. What changed is that a piece of the job stopped depending on my speed at all.

The clearest illustration is a quiet deal. Not a specific one I’ll put a date on — the pattern, which repeats. What brings a quiet deal back isn’t a cleverer message — it’s that the system reconstructs the entire history of a relationship before it says anything, rather than reading the last email and replying to that. A human picking up an old thread reads the top of it, because reading all of it is expensive. The system doesn’t experience that as expensive. So it does the thing I would have done if I had unlimited patience, on a day when I don’t.

I’m not going to give you a number for how much better the documentation got. I have counts, and they’d look good in a paragraph like this one — which is exactly why I’m not using them. What I can tell you is the shape of the change: the record stopped depending on my willpower at 6pm. It gets written the same day, because the same-day version is the only one that’s actually cheap.

I’ll also tell you what I trust least about this. It would be easy to hand the system credit for every good number in my pipeline this year, and I can’t cleanly do that. The market moved. I got better at the job. What I can attribute, specifically, is narrower and more boring: no deal went quiet unnoticed, and every stage change had something behind it.

Leverage isn’t a technology

If this reads as an argument that everyone should go build an AI system, I’ve written it badly.

Software is one form of leverage. It’s not the interesting part — it’s just the form that gets the most attention right now.

A written process is leverage. You decide once how something is handled, and then it’s handled that way without a meeting.

A documentation habit is leverage, which is why the 3-to-17 number matters more than it sounds like it does.

A relationship where the trust is already established is enormous leverage. You skip the twenty minutes of context-building that every new conversation demands, and you get straight to the actual question.

Even a well-made decision is leverage — the kind you make properly once, so you stop re-litigating it at 11pm every few weeks.

The test I use fits in one line: will this hour still be doing work for me in three months?

Most hours won’t. That’s fine — most hours aren’t supposed to. But if the answer is no for every hour of every week, then I’m running an operation that can only ever go as fast as I personally go, and I’ve met the limits of that guy.

The honest cost

Building leverage feels bad while you’re doing it.

It looks like slowness. It produces nothing you can point at on the day. Nobody has ever thanked me for the afternoon I spent writing down how I do something, and there’s a real, slightly humiliating stretch where you’re visibly less productive than the person beside you who is just doing the work.

It also fails in a way pure effort doesn’t. Effort applied to a bad process wastes a day. A system built on a bad process scales the bad process, confidently, at volume, while reporting success — a system will tell you exactly what you configured it to tell you, in a very reassuring tone.

The answer isn’t to abandon productivity. Some weeks are pure throughput, and pretending otherwise is a fantasy. But if every week is throughput, nothing ever gets built, and next year looks like this year with more of it.

What I try to do now is leave one part of each week deliberately unproductive. Not rest — construction. An hour spent making the next hundred hours cheaper.

Most of those attempts don’t work. The ones that do are the only reason my year looks different from my last one.

If you’ve built something like this — a system, a process, a document, anything that quietly kept working after you stopped touching it — I’d genuinely like to hear about it. Especially the part where it looked like a waste of time. Come tell me what you built and what it cost you.