I've noticed two distinct modes in my work. Some days I'm drawn to building something new; other days I'm better suited to steady maintenance. For a long time I didn't ask whether aligning my priorities with that state might lead to better outcomes.
The conventional approach is to prioritize by importance and urgency. If tickets are blocking someone, work the tickets. If the new idea isn't urgent, defer it. Personal energy state is treated as secondary to what's objectively most important.
But state does matter. Ignoring it is expensive.
I learned that recently. I'd woken up with a clear picture of how to make worker restarts graceful, drain logic, SIGTERM handling, the whole thing forming before I'd opened my laptop. Then I spent the morning responding to Slack and fixing scaling issues, they were quick wins. By afternoon the picture was gone. I still wrote the code eventually. It took longer, and the approach was more conservative than what I had in mind at 9am.
Generative mode vs consolidation mode
Creative energy, for me, is when I'm pulled toward a specific problem. I have opinions forming before I've opened my laptop. The design space interests me. I'm willing to sit with hard questions. In that state I generate new ideas, new approaches, things that didn't exist before.
Maintenance energy is capable and consistent, but not generative. I close tickets, fix the deploy script, refactor that module I've been meaning to clean up. I'm executing, not exploring.
The difference isn't interest versus boredom. I'm interested in everything I'm actively building. The difference is whether I'm in generative mode or consolidation mode.
Creative energy is perishable
When I've tried to do creative work in maintenance mode, the output has been mediocre. I push through, I ship something, and it's fine. But fine isn't what creative work should be. I've spent attention without getting the return that state was capable of producing.
The opposite mistake is easier to make and harder to notice: spending creative energy on maintenance work. I'm clearing my inbox, closing easy tickets, telling myself I'm warming up before the real thing. But I've already started, in the wrong direction. I'm burning the best part of my day on work that would've gone just as fast at 4pm.
That worker restart morning was exactly this. Creative energy attaches to specific problems at specific times. If I have it for a problem today and spend it on maintenance instead, the moment is gone. Tomorrow I might have creative energy again, but probably not for that same problem, at that same intensity. The ideas that were live today go quiet. The design question that felt urgent stops pulling. I can return to the problem, but I return to it cold.
Maintenance work doesn't have this problem. The deploy script will still need fixing tomorrow. The tickets will still be there next week. Maintenance work is patient, it waits.
When I have creative energy for something, that's the scarcest input in the whole stack. More scarce than time, at least for me.
Reading the signal
Reading the signal, for me, is simple: am I pulled? If I'm thinking about a problem before I've consciously directed attention toward it, in the shower, on a walk, between other things, that's creative energy. The problem has claim on my mind without me having to decide.
If I'm procrastinating on starting, if the first thing I want to do is clear the easy stuff first, that's maintenance mode. Not bad, just different. The hard cases are when the maintenance work is actually urgent. Someone's blocked. The deploy is broken. I do the maintenance then, but I try to know what I'm trading.
When each mode shows up
Creative energy is most reliable early in the day, before the inbox has loaded, before anyone has asked me for anything. It also shows up after incubation: a problem I slept on is different from a problem I just encountered. Morning creative energy often resolves what I was stuck on the night before, not because I thought harder but because I stopped thinking and let it settle.
It also shows up when I'm underloaded, not overwhelmed. Generative thinking requires cognitive slack. When every cycle is consumed by urgency, there's nothing left for exploration. My best ideas tend to arrive when I'm slightly bored.
Maintenance energy is my default for most of the afternoon, after heavy context-switching, and naturally after a creative push. After I've built something new, consolidation is what's left. I've learned not to fight it.
What I do about it
Protecting mornings has been the main move for creative energy. I keep my phone face-down across the room until I've put an hour into whatever I was planning to build. No Slack, no email, no quick check. The first thing I read in the morning shapes my brain state for the next few hours. If it's someone else's problem, I'm in reactive mode before I've started.
I keep a "parking spot" note at the end of every session, whether I was building or maintaining. One sentence: the exact next thing I'd do if I were sitting down right now. Not "make worker restarts graceful" but "add drain logic in workers/restart.py, starting from the SIGTERM handler." On maintenance work, it makes the next session easier to start. On creative work, it matters more: when I sit down the next morning in good creative state, I don't have to reconstruct context. I just start.
For maintenance, I lower the activation energy and batch when I can. I block a couple of afternoons each week as no-new-work time. The tickets, the PR reviews, the infrastructure stuff, all goes there. When something comes in that isn't urgent, I don't interrupt what I'm building, I add it to the list. I avoided this for years because I thought it meant falling behind. It doesn't. The maintenance gets done in less total time when it's batched.
I don't treat maintenance as the second-best use of my time. The deploy script matters. Maintenance work in maintenance mode is the right allocation.
When I have creative energy for something, I try to work that thing. I don't defer it until after the easy wins. The easy wins don't depend on my state, they go just as fast when I'm in maintenance mode.
Friction helps: a conversation with someone who thinks differently, a book that challenges something I held comfortably, a problem from a different domain that rhymes with mine. Creative energy, for me, sparks from ideas colliding. And when I'm stuck, the move is often to stop and let the problem incubate.
The work I'm proud of almost always came from creative energy well spent. The work I'm embarrassed by came from forcing creative work into maintenance time.
One thing that took me a while to accept: if I have a great creative morning and I'm still in flow at noon, I keep going. The maintenance block can slip. Creative flow doesn't reschedule.