Growing your career
Engineering leverage
Impact divided by time. Ask what the next hour is worth, not whether you looked busy.
Engineering careers are not won by who sits at the keyboard longest. They are won by who repeatedly puts hours into work with outsized leverage — and who can still think clearly in year five.
Concept
Leverage is impact divided by time invested. Two hours spent automating a weekly manual task that takes the team three hours can return hundreds of hours a year. Two hours spent reformatting a doc that nobody reads returns near zero. Effective engineers obsess over the numerator and the denominator.
Edmond Lau's framing beats most productivity advice: do not ask "was I busy?" Ask "what was the highest-leverage use of the next hour?" Sometimes that is deep coding. Sometimes it is a ten-minute message that unblocks five people. Sometimes it is saying no.
Walking through a four-hour window
A simple weekly habit:
- List the outcomes that matter this week (ship X, decide Y, reduce Z).
- Rank tasks by expected leverage, not by urgency theater.
- Schedule the top items into protected focus blocks first.
- Let low-leverage work fill the gaps — or die.
You have four hours free. Options: (A) grind through eight easy code-review nits across random PRs; (B) write the design decision that unblocks a two-week project for three engineers; (C) manually fix twenty rows in a spreadsheet. B is usually correct even if A feels more "done." C should be a script or a ticket to the owning team.
Higher scope is often just consistently choosing higher-leverage problems. Learning is leverage if it changes what you can ship. It is procrastination if it is only courses without application.
| Cadence | Practice | Example |
|---|---|---|
| Daily | 30–60 min on a hard skill tied to the current project | System design notes for the migration you own |
| Weekly | One deliberate practice block | A debugging kata; one chapter; one RFC outside your team |
| Monthly | Ship a learning artifact | Internal talk, doc, small open-source PR |
| Quarterly | One bigger bet | A new domain, or an interview practice cycle |
When job searching, the same idea applies: practice the round you are weak at, not the one that feels comfortable. See interview prep.
Common mistakes
- Measuring busyness. Hours online and Slack responsiveness are not the numerator.
- Choosing work that feels done. Eight nits can lose to one unblocking decision.
- Learning with no artifact. Consumption without a shipped note does not compound.
Check yourself
Four free hours: eight easy review nits, or a design decision that unblocks three people for two weeks. Which?
The design decision. A feels more done; B has the leverage.
You finished three courses this month and shipped nothing from them. Leverage?
Not yet. Learning counts when it changes what you can ship.
Key takeaways
- Leverage is impact per time, not hours or motion.
- Rank the week by outcomes; protect time for the top items first.
- A short unblock for several people often beats a pile of easy tasks.
- Learn on a cadence tied to artifacts, not endless consumption.
Further learning
Edmond Lau at Google on engineering leverage:
Next: protecting focus time.
Related
Productivity, focus and sustainability
Leverage, deep work, saying no, and a pace you can still think clearly at in year five.
Protecting focus time
Hard engineering needs uninterrupted stretches. Treat focus blocks like meetings with a critical stakeholder — you.
Growing your scope
From tasks to features to projects to problems: finding and taking on higher-leverage work.