Fridays, One to Three
Most weeks, the last two hours of my work are spent on something that will never become a paper.
I call it Fridays, one to three. There's usually a colloquium at 3pm, or some other meeting that caps off the week, so 1 to 3 on Friday afternoon is really my wind-down time. By then I've done the things I had to do, and most of the things I should have done, and I let myself spend a couple of hours on something different. Still work. Still research-adjacent. But not research.
The 1% things
I think of these as my "1% things", or side quests. A side quest is a problem that sits next to your research without being your research. It might help your overall goals, or your career, or your community, or your colleagues. But there's no expected paper at the end of it. (Sometimes there are papers, but that's not the goal!) You're not doing it because it will appear on your CV under "publications". You're doing it because it's useful, or interesting, or because someone needed it done and nobody else was going to do it.
Some of my Friday 1-to-3 things have been tools related to NASA/ADS, a shared calendar that doesn't require a human sacrifice to maintain, or a slightly unhinged hack to keep my inbox out of focus, or graphical user interface tools for spectroscopic analysis. None of these were going to be papers. Most of them have been more useful to more people than several of my papers. Some side quests are very research adjacent: learning JAX, reading about a crazy topic that is totally outside of my field. They're still side quests: research-adjacent work that I don't expect to amount to a paper.
The forcing function
The side quest makes the rest of the week better.
Knowing that I get to spend Friday afternoon on something fun is a genuine motivator. It pulls the week forward. It's the reason I actually get to #InboxZero (sometimes!), the reason the boring-but-necessary things get done by Friday lunchtime instead of leaking into the weekend. You can't relax into a side quest if there's a pile of unanswered email staring at you. So the side quest becomes a forcing function for finishing everything else.
"Don't waste your time with that"
Some people will tell you to never do this. They'll say side quests take you away from research, and that every hour not spent on research is an hour your competitors are pulling ahead. They'll say it is a waste of time, and that you risk pigeonholing yourself into someone who isn't serious about research.
They're not wrong. Side quests do take you away from research. If you spend every Friday afternoon building little tools, that's a hundred-odd hours a year you didn't spend on your next paper.
But I think this advice misses something. You need a baseline of high-quality research and sustained productivity to stay competitive and current: that part is non-negotiable, and no amount of side-questing will save you if you neglect it. The side quest is what you do after you've cleared that bar, not instead of clearing it.
And once you've cleared it, ask yourself honestly: what is that extra fraction of research time actually worth to you? Two more hours a week of research, on top of an already-productive week, mostly buys you a slightly faster version of what you were already doing. The marginal return is real but small. I'd argue that the same two hours spent on the right side quest can return far more.
The networking nobody tells you about
The clearest evidence I have is this: in my experience, far more people know me for my Friday 1-to-3 things than for my research.
That sounds like it should be a problem. It isn't. People discover these things and they know who you are before they know what you research. Is that a problem? I don't think so! In a crowded field I think any differential networking is beneficial to you, especially if it means you are expanding your network to people outside of your immediate research field. You end up talking to people you would never have met through your research, in fields adjacent to yours, at institutions you'd never visit. A surprising number of collaborations, invitations, and opportunities have started with someone saying "oh, you're the person who made that thing".
That networking has real value attached to it. It compounds. And you generally can't manufacture it on purpose — it's a side effect of building something useful and letting people find it. The extra two hours of research almost never produces that; the side quest does it routinely.
Do the side quest, but mind the primary quest
So: do the side quests. Give yourself the couple of hours. Build the tool, fix the thing, help the colleague, chase the rabbit down the hole.
But keep the order straight. The side quest is the dessert, not the meal. It only works because the primary quest comes first — the research that keeps you competitive, the obligations you've actually met. The moment the side quest starts deterring you from the primary quest, it stops being a 1% thing and starts being a problem.
Get your other stuff done and keep your research strong. Then, on Friday, from one to three, go do something that will never be a paper.