The Final Twenty Percent

Most mornings I do the same thing. I open a notebook and write down everything I’m carrying around in my head, which on a normal day is thirty-odd items and on a bad one is more, and I don’t filter or prioritize while I’m doing it, I just keep going until there’s nothing left rattling around.

Then I go back through and put a little acronym next to each one so the chaos sorts itself into buckets: L for life, BB for the B2B product I’ve been building for years, SA for my wife’s company, JOB for the day job, and a few others depending on the season. The acronyms don’t change anything about the tasks, they just make the page look like it belongs to someone who has things under control. Then I read the list one more time, slower, cross-reference it against yesterday’s to make sure nothing got dropped, and circle the three to five things I’m going to do today.

It feels productive, and by any reasonable definition it is a productivity practice, and for years I assumed it was working. It was working, just not at the thing I thought it was working at.

What I eventually noticed, after enough mornings of this, was that the things I circled weren’t the things that needed doing. They were the things I wanted to do, which is a different list, and the gap between those two lists turned out to be more or less the whole story.

I have ADHD, and the shape it takes in me is that I love the complicated part of any job. The architecture, the interesting decision, the first eighty percent where the problem is still fresh and the solution could go anywhere. I’m good at that part, and I’ll happily start it at seven in the morning. The last twenty percent is a different matter. That’s the proofread, the follow-up email, the deployment checklist, the phone call I’ve been putting off, the form, the receipt, the part where the interesting work is over and what’s left is administrative or repetitive or slightly awkward, and the only thing between me and done is a willingness to grind through something that isn’t fun.

So when I sit there circling my three to five in the morning, what I’m actually doing, without quite admitting it, is picking the ones where I get to start something. I’m using the ritual to put off the final twenty percent of yesterday’s work, and the day before’s, and the week before’s, and those items drift to the bottom of the page where I copy them out again the next day and circle something else instead.

None of this is the ritual failing. The ritual does exactly what it was built to do, which is make me feel like I’m on top of my work. It was never built to make me finish anything, and it took me an embarrassingly long time to notice the difference.

I don’t think this is only a me problem, either, because every task app I’ve ever used is built on the same assumption my notebook is: that the hard part is getting everything out of your head and into a system, and once it’s captured and filed and tagged, the doing will follow. That’s the promise behind GTD and second brains and every app with infinite folders and projects and views, and I bought it for a long time.

For me it was backwards. Capture was the easy part. Capture was the bit I did every morning that made me feel productive. The hard part was resolution, which is the moment where you look at something you’ve been writing down since March and either do it or admit out loud that you’re not going to, and delete it. Better capture didn’t help with that at all. If anything it made things worse, because it lowered the cost of adding new things while doing nothing about the cost of finishing them, and every well-organized list I ever kept eventually turned into a very tidy graveyard. The folders gave the rot somewhere respectable to live.

I think the reason capture feels like progress is that it produces something you can see. The item that was bouncing around in your head is now in a system, and you feel lighter, and for a second that feeling is almost indistinguishable from the one you’d get from actually finishing the thing. It’s real, and it’s not nothing, it just isn’t what I thought it was. I’d moved the task sideways, out of my head and into a queue I wasn’t going to look at, and because the relief was real the whole system quietly trained me to do the same thing again tomorrow.

What finally broke the loop was being honest about two questions I’d been avoiding.

The first was how many things I can actually finish in a week. Not start, not make progress on, not get eighty percent of the way through, but finish, cross off, never think about again. The honest number was small, somewhere between five and ten depending on what they were, which is embarrassing next to a list that ran to thirty items every morning.

The second was how many of the things on the list I was actually going to do. Not would like to do in principle, or felt I should do, but going to do. That number was also small, and a lot of what was left over had stopped being relevant months earlier and was only still there because I kept copying it forward out of habit.

If you take both of those numbers seriously, the conclusion is a bit uncomfortable, which is that most of the list was fiction. The morning ritual wasn’t planning my day so much as reassuring me that I was the sort of person who had his obligations under control, while the actual obligations aged quietly in the margins until I either did them or admitted I wouldn’t.

So I started wondering what a list would look like if it wasn’t allowed to lie to me. It would have to be short, probably very short, so that I’d have to choose what I was carrying instead of just carrying everything. It would have to make the old stuff impossible to skim past. And it would have to treat not doing something as a real decision with a real button, instead of letting things slide off into that ambient state where they’re not done and not quite not-done either.

That’s the app I ended up building, and I built it for myself before I had any plan to show it to anyone.

It has a hard ceiling on how many tasks it will hold, a handful, few enough that adding something means deciding what it’s replacing. Every task shows how long it’s been sitting there, and the list is sorted by that, so the thing I’ve been avoiding for forty-seven days is at the top every time I open it and there’s no folder I can move it into to make it go away. And when I’m done with something there are two buttons: Complete, which means I did it, and Close, which means I’m letting it go on purpose. A completed task goes to a Resolved page with its milestones, so if I was wrong about being done I can pull it back without starting over, and a closed one is just gone, which is the point of closing it.

That’s more or less the whole thing. The notebook still happens every morning, because emptying my head is still useful, but the list I actually act on lives somewhere else now, short and aging in plain sight. Most days I tap Complete. Some days, more than I’d like, I tap Close, and I’ve stopped feeling bad about that, because closing something on purpose turns out to be a lot healthier than rewriting it every morning for a year.

The app then sat in TestFlight for a couple of months, because releasing a to-do app whose main feature is not letting you add to-dos seemed a bit silly, and also because releasing it was, of course, one of the things I was avoiding. It’s called Unresolved. I did see the irony eventually.