StoweA productivity system stops working when the thing carrying it was novelty, because novelty always runs out. What is left is whatever the system costs to use — and most systems cost more than they are worth on an ordinary bad day. The collapse is not a verdict on your discipline. It is what happens when the interest expires and the underlying economics show through.
Two weeks is the figure people name for how long it takes. I have no measurement behind it and neither does anyone quoting it; what I have is the same pattern in my own abandoned systems, several times over. The useful part is not the number but what happens at the end of it, and I think three structural properties decide that: what the system charges you daily, what it does when you skip a day, and whether it pays you anything for partial use.
Why does the first two weeks always go so well?
Because a new system is interesting, and interest is doing the work that habit has not had time to build. Setting it up is itself enjoyable — choosing the notebook, naming the categories, building the template — and that enjoyment is easy to misread as evidence that the system fits. It is evidence that setting it up was fun.
The test of a system is not the first fortnight. It is the first genuinely bad week: overslept, three things on fire, no appetite for maintaining anything. Every system looks good under ideal conditions. Only some are usable under real ones, and that is the only comparison that predicts whether you will still be using it in March.
What does a system charge you every day?
Most systems charge rent in the form of decisions at capture time, and that rent is what makes them unaffordable on a bad day. If adding a thought requires choosing a project, a priority, a due date, an energy level and three tags, then capturing costs maybe forty seconds and, worse, five small decisions. On a good day you pay it. On a bad day you skip it — and one skipped capture is one thing the system no longer knows, which makes the system slightly less trustworthy, which makes checking it slightly less rewarding.
A low-rent system lets a thought in immediately and sorts later, or not at all. Cheap capture is also the first of the three properties that predict sticking in paper versus digital planners; what that post treats as a medium-independent property, this one treats as the rent a system charges. It is the same argument behind tags versus folders: a folder demands a filing decision at the worst moment, which is exactly the decision an overwhelmed brain defers. Anything you can do after capture should not be required at capture.
What happens when you skip a day?
In most systems, a skipped day creates a chore: a reconciliation you have to complete before the system is usable again. That chore is what turns a gap into an ending. A bullet journal needs migrating, a calendar of blocks has turned false, a list of overdue items has gone red — and each of those is aversive, so it is deferred, so the backlog grows, so it becomes more aversive. Systems rarely die at the skip. They die at the reconciliation.
The property worth demanding is that returning after a gap costs nothing: the plan for today is just today, and what was parked is still parked, not accusing you. This is why I built Stowe's run card to count days planned rather than days missed, with nothing red and no streak to lose — described in how Stowe keeps a run without the shame. The general principle holds on paper too: if a missed day produces work, the work is the thing that will end your system.
Does partial use still pay you anything?
A system that only pays out when fully maintained will eventually pay out nothing, because nothing stays fully maintained. This is the quiet killer of elaborate setups — the review cadence, the cross-linked databases, the weekly cleanup. They deliver real value at full maintenance and fall off a cliff below it, which means every imperfect week is a total loss rather than a partial one.
The alternative is a system where each piece is independently useful. Capture pays you the moment a thought is out of your head, whether or not you ever sort it — that is cognitive offloading, and its value is well established (Risko & Gilbert, 2016). Picking three things for today pays off even if yesterday's list was never reviewed. When the parts pay independently, a bad week costs you a bad week rather than the whole system.
What three questions should you screen a system on?
Screen on cost at capture, cost of a gap, and value from partial use — in that order, and asked about a bad week rather than a good one:
- What does capture cost when I am late and overwhelmed? If the answer involves more than a few seconds or any decision that could be made later, it will be skipped exactly when it matters most.
- What does the system look like after I skip three days? If the answer is a backlog, a red list, or a broken streak, the gap will become the ending. It should look like nothing happened.
- What do I get from using half of it? If half the system is worth nothing, every imperfect week will be a write-off.
Stowe is my answer to those three, so treat my endorsement accordingly: talk for ten seconds and the thought is in, the day is 3–5 things, a gap leaves nothing red, and capture alone is useful even in a week you never plan. But the questions matter more than the answer. A paper notebook that scores well on all three will outlast an elaborate app that scores well on none — and if you have been through several systems already, that is not a discipline problem. That is a screening problem, and screening is fixable.
Frequently asked questions
Why do I abandon every productivity system I start?
Usually because the system was carried by novelty rather than by being easy, and novelty is a fuel that always runs out. Three structural properties decide what happens next: how much daily maintenance the system charges, how gracefully it survives a skipped day, and whether using it partially still returns value. A system that needs daily upkeep, looks broken after a gap, and only works fully-maintained will end — not because of your discipline but because that is what those properties produce.
How do I stick with a planning system long term?
Choose for the bad week, not the good one. Ask what using the system costs on a day when you are late and overwhelmed: if capture takes seconds and requires no filing decision, it survives; if it needs a category, a priority and a project before an item can exist, it does not. Then make sure that coming back after a gap costs nothing — no backlog to reconcile, no broken streak, nothing red.
Is the two-week collapse an ADHD thing?
It happens to everyone, and it happens faster with ADHD. The novelty of a new system is genuinely motivating, and when interest-driven motivation is a larger share of what gets you started, its fading takes more with it. The fixes are the same for everyone, just less optional: lower maintenance, no punishment for gaps, and value from partial use.
Should I just try a simpler system?
Simpler helps, but "simple" is not quite the right axis — a simple system can still charge daily rent, and an elaborate one can be forgiving. The properties that matter are cost at capture, cost of a gap, and value from partial use. Judge candidates on those three rather than on how minimal they look.