The Stop-Doing List: Why Deciding What to Drop Beats Any To-Do System

My to-do list has never once been empty. Not in twenty years of keeping one. I have changed systems maybe nine times — index cards, a wall of sticky notes, three different apps, a plain text file, a bullet journal I kept faithfully for about six weeks — and the one constant underneath all of them is that the list on Friday is longer than the list on Monday.

For a long time I read that as a tooling problem. If the list keeps growing, the capture must be wrong, or the prioritisation, or the weekly review I keep skipping. So I would go shopping for a better system, which is its own species of procrastination, and the new system would work beautifully for about eleven days.

What finally landed is that nothing enters my life through the to-do list. Things enter through conversations. A yes said quickly in a hallway. A yes typed into a thread because typing no would have taken four more sentences. A yes given to myself at eleven at night about a side project I will not remember caring about by Thursday. The list is only the ledger. Reorganising the ledger does nothing about the spending.

The arithmetic of yes

A yes is cheap at the moment you say it and expensive on the day it comes due. That gap — sometimes weeks wide — is where nearly all overwhelm lives. You are not, mostly, drowning in work you agreed to this morning. You are drowning in work agreed to by a version of you who had no idea what this particular week would look like and who, frankly, was not thinking very hard.

Worse: commitments do not expire on their own. A standing meeting created in March for a project that shipped in June is still on the calendar in October, and will be there next year, because calendars have no garbage collection. Nobody schedules a meeting to end a meeting.

Any engineer recognises this shape. Dependencies get added in a rush and removed never. Feature flags outlive the experiments they were guarding. Someone writes a cron job, leaves the company, and the job runs for six more years doing something nobody can quite describe. Starting has a ceremony — a ticket, a kickoff, a name. Stopping has no ceremony at all, which means the drift only ever runs one direction.

Your week works the same way. And a to-do app, however good, is a tool for arranging what is already inside the system. It has no opinion about the door.

Why no is harder to operationalise than yes

Here is the asymmetry that makes this so stubborn: yes is one action, and no is a negotiation.

Yes ends the conversation. It is a single word, it makes the other person happy, and it costs you nothing you can feel in the room. No opens a conversation. It requires a reason. And every reason you offer is a surface for counter-argument — say you are busy and you will hear that it is only an hour, say the timing is bad and you will be offered a different month, say you are not the right person and you will be told you are exactly the right person. The refusal has to survive a round of rebuttal that the acceptance never faces.

There is also a timing problem. The cost of saying no lands immediately, in someone's face, in a slightly changed atmosphere. The benefit arrives months later, diffusely, as work that never showed up. You cannot feel an absence. So the ledger is rigged: a vivid present cost weighed against an invisible future gain, judged by a nervous system that heavily discounts anything further away than next Tuesday.

And underneath both: yes feels like it is about the task, and no feels like it is about the person. That is the part that actually stops people. Not laziness, not weak boundaries in the abstract, but the entirely correct perception that a decline made in the moment reads, to a human being standing in front of you, as a small judgment about them.

What writing it down changes

A decision made on the spot is personal. A rule made in advance is not.

That is the whole mechanism, and it is worth sitting with, because it sounds like a technicality and it isn't. When I decline something in the moment, I am making a choice about you, now, in front of you. When I decline something because I wrote three months ago that I no longer take calls before ten, I am reporting a fact about how my life is arranged. The first is a verdict. The second is closer to a weather report, and nobody argues with a weather report.

There is real evidence for this beyond intuition. Vanessa Patrick and Henrik Hagtvedt published a study in the Journal of Consumer Research in 2012 comparing two ways of refusing a temptation: "I can't" and "I don't." Participants who framed refusals as I don't do that held to their goals far more often than those who said I can't. The wording is doing something structural. "I can't" points at an external constraint you are half apologising for and half inviting someone to argue with. "I don't" points at who you are, which is not a debatable position.

A written stop-doing list is a machine for manufacturing "I don't." You do the hard part once, at a desk, on a quiet Sunday, with nobody's face in front of you. Then you spend the following month reporting a decision rather than making one, eleven separate times, under pressure.

Jim Collins made a version of this argument in Good to Great, and the framing has stayed with me: the stop-doing list matters more than the to-do list, because prioritisation without a stopping mechanism isn't really prioritisation. If capacity is treated as unlimited, ranking is decorative. You have not chosen your top three until something has actually been dropped to make room for them.

What actually ends up on it

When I started keeping one, I expected it to fill with obvious time-wasters. It did not. Almost nothing on my list was stupid. That is precisely why it had survived.

Things that were true once. The largest category by far. The recurring sync for a project that shipped. The newsletter subscribed to for a job left two years ago. The spreadsheet maintained for a decision already made. None of these were mistakes. They were correct commitments that quietly outlived their reason, and nothing in the system is designed to notice that.

Things kept because you are good at them. The competence trap. You are the person who fixes the deploy scripts, so you get the deploy scripts, forever, including on the weeks you would rather be doing the work you were actually hired for. Skill attracts requests. Left alone, it will fill your calendar with your second-best work.

Things done for an audience that is not watching. The polish that goes into an internal document three people will skim. The lengthy reply written to a message that needed one line. Some of this is genuine care and worth keeping. Some of it is a performance for a spectator you invented.

Ambient availability. Not a task at all, but a standing, unlimited yes. Notifications on. Reachable at all hours. Open to being interrupted by anything, by anyone, at any time. This single item costs more than most of the others combined, and it never appears on a to-do list because it has no name.

A monthly stop-doing list that survives contact

Fifteen minutes, last Sunday of the month, three lines. Three. The temptation is to write twelve and keep none.

Each line takes the same shape — the thing, and the sentence you will actually say when asked:

I no longer ______. If asked, I say: "______."

Writing the second half is not optional, and it is the part everyone skips. An unstated no collapses the moment someone pushes, because you are composing under pressure. A rehearsed sentence is available instantly and comes out level.

Three starter categories, one line each:

1. A commitment. Something recurring you have agreed to — a meeting, a role, a favour that hardened into a standing arrangement. I no longer attend the Thursday status call. If asked, I say: "I'll read the notes. Pull me in directly if you need a decision."

2. An input. Something that consumes attention rather than time — a feed, a newsletter, a group chat, a category of news. I no longer check the news before the day starts. If asked, I say: "I catch up in the evening."

3. A standard. A quality bar you hold yourself to that nobody asked for and nobody would notice you lowering. I no longer format internal documents. If asked, I say: "It's a working draft."

That third category is the uncomfortable one and usually the most valuable. A great deal of what exhausts people is not obligation arriving from outside. It is a private standard, adopted years ago for reasons that made sense then, now enforced by nobody but you.

Reviewing without re-litigating

Give every entry ninety days and one question at the end: did anything actually break?

Almost nothing breaks. That is the finding, and it is the whole reason to keep a written list rather than simply make the decisions. The meeting you stopped attending carried on without you and the project shipped anyway. Nobody mentioned the unformatted document. The world's indifference to your withdrawal is mildly insulting and enormously freeing, and you will only ever learn it by running the experiment and writing down what happened.

Two rules keep the review honest. Don't delete old entries; move them to the bottom under a heading and leave them there. The pattern of what you keep re-adding, month after month, tells you something no amount of reflection will — those are the yeses you have not really decided about yet.

And don't use the review to re-argue an entry. If something genuinely needs to come back, let it come back through the front door, as a new yes, with the same scrutiny anything new would get. The point of writing a decision down is to stop paying for it a second time.

I keep mine in the same file as everything else, which means I see it during morning practice more often than I would like. There is a version of stopping that is really avoidance wearing productivity language, and having the list somewhere you cannot skip past is a reasonable guard against that. The question I try to hold is not what can I get out of but what am I doing out of momentum rather than choice. Those two look identical from the outside and could not be more different from the inside.

Questions worth asking

Isn't a stop-doing list just quitting with better branding?
Sometimes, and worth checking honestly. A useful test: does the thing get harder or easier as you do it? Quitting tends to target what is difficult. A good stop-doing list mostly targets what is easy — familiar, mildly pleasant, and no longer connected to anything you want. The entries that make you uncomfortable are usually the correct ones.

What if I can't stop? My manager assigned it. My family expects it.
Then it doesn't belong on the list, and putting it there only makes the list a place to feel bad. The list is for what sits inside your control. What often is inside your control is the standard rather than the obligation: you may not be able to stop producing the weekly report, but you may well be able to stop making it beautiful.

Three feels like too few.
It's deliberate. A list of twelve is a wish. Three fit in your head, which matters enormously, because the value of this whole exercise is being able to recall the rule at the moment you are asked. If you can't say your list from memory in the middle of a conversation, it will not help you in the middle of a conversation.

How do I say no to someone I don't want to disappoint?
Give the decision, not the justification. A reason invites negotiation; a fact does not. "I'm not taking new projects this quarter" ends somewhere. "I'm pretty slammed right now" is an opening bid. And it helps to remember the shape of the trade: the disappointment is real but small and short, while the resentment of an obligation you never wanted is real and long.

How long before this stops feeling awful?
Longer than you want. The first few declines feel wildly disproportionate to what they are. Somewhere around the third month, when nothing has collapsed, the whole thing starts to feel less like transgression and more like ordinary maintenance.