Your Notes App Isn't the Problem. Your Filter Is.

Some people have solved capture completely and are drowning anyway.

The notes app is on their home screen. They photograph whiteboards, carry a notebook, and keep index cards by the keyboard, so nothing gets lost: the article worth reading, the half-idea from the commute, the "we should look into that" from a Tuesday meeting. Two years later all of it is still there. Opening the app feels bad, so they open it less, and the pile grows faster.

The standard advice is to switch apps or adopt a framework. Neither helps, because capture was working all along. What's broken is the exit: things go in and never come out.

Why good judgment builds a bad pile#

Ask someone why they saved an item and you'll hear some version of "it was interesting" or "it might come in handy." They're usually right, which is exactly why the pile is so hard to cut.

Each item gets judged on its own: is this worth something? Nearly everything passes. Someone who saves a hundred items a month and can defend every one of them has good taste. They also have a hundred things to do and time for maybe ten.

A better question compares: is this worth more than whatever it would push out? Very little survives it, and that's the intent. A filter that measures value lets everything through, because value is cheap and attention is what runs out.

The second problem is that the pile mixes kinds of things. A saved link and a promise to a colleague share only the fact that both got written down. Skip the link and nothing happens; skip the promise and your colleague loses a week. Kept in one list, they add up to a single debt that looks unpayable, and opening the list produces dread instead of action.

Everything below fixes one of those two problems: the filter that judges items alone, or the list that mixes them.

Declare bankruptcy on the backlog#

Nobody works through a two-year backlog. The attempt goes the same way every time: a Sunday afternoon, forty items in, three forgotten ideas rediscovered and added back, and a longer list at the end than at the start.

Skip the processing. Move everything older than about sixty days into an archive where you won't see it. Archiving feels safer than deleting, even though you will rarely open either. Don't read the items on the way out. Reading turns a five-minute job into three weeks, and every item you read gives loss aversion another chance to keep it.

Put a reminder in your calendar three months out with one question on it: how many times did I go back to the archive? Most people get to once or twice. That count tells you what the hoard was really worth, and it makes the next round of archiving easier.

Items younger than sixty days get triaged in short, timed passes: twenty-five minutes, about five seconds per item, and anything that needs longer goes to the archive by default. The timer exists to cut off deliberation, since deliberating over each item is how the pile got built.

Four destinations, and only four#

Sorting a hundred items by priority takes an afternoon and leaves all hundred where they were. Sorting them by kind decides what happens to each one.

Reference is where most people go wrong. They treat information as a task, so a saved pricing page sits in the list for years, faintly asking to be dealt with. Put reference material somewhere searchable and stop looking at it. Search will find it when you need it.

The table has no "someday/maybe" row, and no "on deck" either. Lists like that keep the item and the guilt together, and you reread them every week without acting on any of it. Keeping the item outright or dropping it would both cost less.

Two destinations come back every week. The other two never come back on their own.

Ten items, sorted#

A typical working week leaves something like this behind.

Out of ten: one commitment, one project, one thing done on the spot, two filed as reference, and five gone. A healthy pass looks roughly like that. If more than a third of what you review survives, the filter isn't cutting.

Good ideas come back#

Junk is easy to drop, and nobody agonizes over a link saved by accident. The items that hurt are the promising ones.

The bet that makes it bearable is simple: if an idea is any good, it comes back. You think of it again in the shower, or halfway through a conversation, or while staring at the exact problem it would solve. When it returns that way, it brings the problem with it. Compare that with a two-line note in a Sunday review, where you first have to work out why you wrote it down.

Recurrence filters better than judgment does, and it runs for free. Judgment asks whether an idea beats nothing, and everything beats nothing. Recurrence asks whether the idea can win your attention against everything else going on. Three unprompted returns are evidence of pull. One note in an app proves only that you had the thought once.

So the archive is where ideas wait until they prove themselves by coming back. It does that job even if you never open it. Browse it on purpose, though, and it turns back into a to-do list.

Some good ideas will be lost for good, and that's the price. In return you get back roughly fifteen hours a year, which is what it takes to reread two hundred items at five seconds each, every week, without acting on any of them.

The twenty-minute weekly review#

Capture happens all day and needs no decisions. Deciding happens once a week, for twenty minutes, at a fixed time. Which time you pick matters much less than keeping it fixed, because a review without a slot slips by a day, then a week, then stops happening.

Most of the twenty minutes goes to sorting, not reading.

Spend the first five minutes just consolidating. Photograph the whiteboard, type up the sticky notes, flip through the notebook, and get everything into the one inbox without judging any of it. You'll want to start deciding here, and if you do, five minutes becomes forty.

Use the next three minutes on anything that takes less than two: send the message, ask the question, cancel the subscription. That usually clears fifteen to twenty percent of the inbox, and it's the only part of the review that feels good while you're doing it. Putting it early gives you momentum for the rest.

Sorting gets ten minutes, at about five seconds per item. Hesitation is your signal: the moment you catch yourself thinking "well, it depends on whether," you have your answer, and it's the archive.

The last two minutes are an audit. If there are more than five active projects, one goes to the archive. Marking it "paused" doesn't count, because a paused project still takes up space in your head while producing nothing. Then read down the commitments and ask which ones you won't actually do.

After that, close the inbox until next week.

Order commitments by who is blocked#

Most commitments are small and similar in size, so ranking them by importance is busywork. Two questions decide the order.

First, is anyone blocked? When a person can't move until you act, that item goes to the top no matter how small it is. A five-minute approval holding up three people beats a two-hour task nobody is waiting on. Your own list won't show this, because it records how much work each item is for you and says nothing about who's waiting.

Second, is the deadline real? A hard deadline has a consequence you don't control: a contract lapses, a meeting happens, a filing window closes. A soft one is a date somebody picked because picking dates feels responsible. Much of the weight of a long commitment list comes from treating soft dates like hard ones.

Blocked people first, then hard deadlines, then everything else roughly by date.

The step almost everyone skips is renegotiation. In each weekly pass, find the commitments you won't keep and tell the person now, while the due date is still weeks away.

"I'm not going to get to the vendor review by the 30th. Can it move to the 12th, or should it go to someone else?"

Sent on day three, that message costs almost nothing. Sent on day twenty-nine, it wrecks someone's plan and costs you some trust. Early renegotiation is an ordinary professional move, and treating it as one keeps the list honest.

A list that's overloaded all the time has a different cause. You're agreeing to requests you would have archived if the idea had been your own. The fix comes earlier, at the moment someone asks, when declining takes one awkward sentence.

Pick the next project#

When a project finishes, leave the slot empty for a few days. Fill it right away and the newest or loudest idea wins by default. After a short wait you'll usually notice one option that keeps coming back on its own.

Then put five or six candidates through these questions:

• Can you describe "done" in one sentence? "Improve our onboarding" is a topic; "a new onboarding checklist used for the next three signups" is a project. About half of a typical list fails here.
• Is something forcing the timing, like a hiring wave, a renewal date, a broken system, or a colleague who just freed up? Without a trigger, projects tend to stall around week three, once the early excitement wears off.
• Does anything else depend on it? A project that unblocks two or three others is worth more than its own payoff suggests.
• Is a window closing? If one option gets much harder in three months and the other doesn't, start with the first.
• Do you actually want to do it? A project you mildly dread runs to three times its estimate and drags everything else down while it lasts.

For a tie, ask which one you'd be more annoyed to find still undone in six months. Enthusiasm tells you how you feel this week, and imagined regret is a better guide to what you care about.

Four reasons sound convincing and shouldn't be. "It's been waiting longest" means it has failed to pull you for a long time. "I've already started" is sunk cost, so finish it now or archive it properly. "It's the easiest" works once, for momentum, but as a habit it keeps you away from the project that matters. And "it's the newest" is the strongest pull in the pile and the least reliable.

Before you start, write the done sentence into the project entry. Without it, finished projects never quite close. They linger as vague ongoing work, and the active count drifts past five without anyone deciding it should.

Tests with scarcity built in#

A popular test asks, "Would I do this next week if my calendar were completely free?" Some people answer yes every time, and they mean it. A free calendar means unlimited time, and with unlimited time nearly everything is worth doing. Each test below makes you give something up to say yes.

None of these asks whether an item is worth doing, since the answer is nearly always yes. They ask what it costs and what it would push out.

Build it in Notion#

Good rules tend to die inside a messy setup, so here is a specific one. It uses Notion, and the same shape works in most tools.

One database, not four#

Most people build separate databases for tasks, ideas, reading, and reference. That splits up exactly what you're trying to gather in one place. Worse, it forces a category choice at the moment of capture, when you know the least about the item. Use one database and let everything land there.

Seven properties#

Every property is one more decision at capture or review, so keep the list short.

Leave out Priority, Energy, Context, and Area. They give you something to adjust without deciding anything, and none of them changes what you do next.

One optional property pays for itself: a Resurfaced counter that you bump each time an archived idea comes back unprompted. It turns recurrence into a number you can see. An idea that reaches three has earned a project slot on its record.

Three views#

Inbox shows Type is Inbox and Created is within the past 30 days, oldest first. That date filter does more work than anything else in the setup. Other systems make you decide to let an item go, which is the one decision this kind of person hasn't managed in years. Here, unprocessed items drop out of view after thirty days with no deletion and no decision, and search still finds them. Start at thirty days and tighten to fourteen if the inbox still feels heavy after a month. If you catch yourself widening the window to ninety days or more, the pile is coming back.

Now shows commitments and active projects grouped by type, blocked items first and then by hard date. It should fit on one screen. If you have to scroll, you have too many commitments or too many projects, and renegotiation or eviction will fix that where a cleverer sort order can't.

Search has no filter and includes the archive. You only reach it by typing a query, which is how reference material is meant to be used.

Don't build a view for browsing the archive. If one exists you'll scroll through it, and scrolling brings back everything you just let go.

Capture and templates#

Capturing an item should take no decisions. On the phone, use a home-screen widget or a share-sheet shortcut. Use the web clipper for links, expecting to delete most of them at review; it's there for the fifth worth keeping. During the weekly consolidation, photograph each whiteboard or notebook spread as a single item. Type a whiteboard up as eight entries and one meeting becomes eight obligations.

Templates hold the structure in place. The Project template sets the type and puts the done field at the top, so a project you can't describe in a sentence gets caught before you spend time on it. The Commitment template asks who asked and whether they're blocked. Inbox items get no template, because capture has to stay instant.

How the system breaks#

The failures look different and end in the same place, with every item judged on its own merits again. Within a month or so a fifth category tends to appear, "quick wins" or "maybe soon," which is the old pile with a new label. Sometimes the setup itself becomes the project: new properties, a dashboard, a different app. If you're redesigning it more than once a quarter, the redesign is how you're avoiding the review. Exceptions are the third route. One exception for a great idea is fine, but after the second, every rule is up for negotiation. And when the Now view outgrows one screen, the cap has stopped forcing the comparison it was there to force.

Start with the archive#

The cheapest first step is the sixty-day archive. Move the old pile out of sight this week, without reading it, and set the three-month reminder. The four destinations, the weekly review, and the Notion views are all easier to set up once the backlog has stopped staring at you.

After that, the system runs on two habits. Everything you capture goes into one place with no sorting, and once a week you spend twenty minutes deciding where each item belongs, with five active projects at most and an inbox that forgets anything older than thirty days. When a new idea arrives between reviews, write it down and let it wait. If it was worth doing, you'll find yourself thinking about it again soon enough, and the next review can weigh it against what it would push out.