Twenty Minutes a Day for Technical Breadth
Picking Up Where I Left Off
In my last post, Why Technical Breadth Matters More Than Ever, I argued that breadth is the new leverage. In a world where AI makes depth cheap, the engineers who know what exists win more often than the ones who know any single thing the deepest. I ended with a habit I’d been quietly running for a while: 20 minutes a day, learning a specific topic.
The idea isn’t mine. I picked it up from the same book I leaned on last time, Fundamentals of Software Architecture by Mark Richards and Neal Ford. They call it the 20-minute rule: set aside at least 20 minutes a day to learn something new or go deeper on a topic. The point is to move things out of “stuff you don’t know you don’t know” and into “stuff you know you don’t know”, the tier where breadth actually lives.
How I Pick the Topic
I keep a running list of things I want to study. Nothing fancy, just a note in Obsidian I add to whenever I hit something I don’t really understand.
The list grows from a few places. Sometimes it’s a gap at work, a tool or service I’ve been using without really getting it. Sometimes it’s a term I keep seeing that I’ve been nodding along to instead of actually learning. And sometimes it’s whatever’s in front of me that day, a new service I need to wire up, or a decision I need more context to make.
That’s the whole process. No curriculum, no study plan. Just the gap that feels most worth closing that morning.
What 20 Minutes Actually Looks Like
Once I’ve picked the topic, I don’t go digging through docs or a 40-minute YouTube video. I ask Claude to write a simplified markdown about the topic, with a concrete example I can follow.
That prompt does a few things at once. It keeps the explanation short and structured, not a wall of text. It strips out the jargon I’d otherwise have to google on the side. And the example ties the idea to something real, usually code or a small case I can work through.
A few topics I’ve been through in the last few weeks:
- Event-driven vs request-driven architecture: when each one actually makes sense, and what breaks when you pick the wrong one.
- Kafka: what problem it really solves, and how it differs from a plain message queue.
- DNS: filling in a gap I’ve had for years on what actually happens between typing a URL and getting a response.
None of this makes me an expert. That’s not the point. Twenty minutes is enough to move a topic from “I’ve heard of it” to “I get the shape of it, and I know when it applies.” That’s the second tier from the pyramid I wrote about last time. That’s the whole goal.
The Forgetting Problem
Here’s the part nobody talks about when they talk about learning a little every day.
You forget.
If I spend 20 minutes on Kafka on a Tuesday and never touch it again, by the next month most of it is gone. The shape is still there, vaguely, but the details I’d need in a real conversation have slipped away.
The fix, for me, is spaced repetition. After each session, I drop a handful of cards into Anki: a question on one side, the short answer on the other. Then Anki feeds the cards back to me on its own schedule. Cards I get right show up less often. Cards I blank on come back soon.
Why It Compounds
Twenty minutes doesn’t feel like much on a given day. That’s part of why it works. It’s too small to skip and too small to burn me out.
But the effect adds up quietly.
The clearest payoff, honestly, is how it’s changed the way I work with AI. The more topics I’ve been through, the better my prompts get. I can name the pattern I’m reaching for. I can ask “would this be a better fit as event-driven?” instead of “how do I make this faster?” AI is great at going deep on a topic once you point it at the right one. Breadth is what tells me which one is the right one.
It also shows up in conversations. In design reviews, in architecture calls, in the moment where someone on the team asks “what do you think about using X?”, I’m more often able to have an opinion, or at least a useful question. Not because I’ve mastered X, but because I’ve spent 20 minutes on it once and the shape is still there.
Stretch that out a few years, and the map gets wide. The toolbox gets big enough to reach for the right thing. The conversations I can drive go up in quality. That’s the leverage I’m really after.
Keeping It Simple
That’s really the whole habit. 20 minutes a day. A coffee. A simple prompt. A few Anki cards.
It doesn’t feel like much on any given morning. But over weeks and months, the map keeps getting wider. That’s the kind of slow compounding I care about most these days.
✌️