✦ · Why it works
The research behind Seiza
Seiza should help because of research that has held up, not because it sounds like it should. Here's what each feature rests on, how strong the evidence is, and what's been left out because it didn't hold up.
- Strong
- A large meta-analysis, review or field experiment.
- Modest
- Real, but small, or a single study.
- Reasoning
- Sensible, but not tested.
01 · What Seiza does, and why
Each feature, and what it rests on
Modest · one study
A pause and an easy way out cut how often you give in
In a six-week field study of the one sec app, people opened the apps they chose 57% less. The option to back out mattered most and the wait helped; the message on its own did nothing.
In Seiza The redirect page: five minutes of focus first, and Not now always right there.
Strong · a large field experiment
Limits you set yourself work
Letting people cap their own future screen time cut their social media use a lot, and short-term help to cut back kept working after it ended.
In Seiza Profiles, Hard mode and its passes, and Restricted sites.
Strong · a review
Habits are set off by their surroundings
Habits form from repeating something in the same context, and after that the context itself sets them off. Taking the cue away works better than fighting it every time.
In Seiza Taming a site: its feeds, Shorts and recommendations are hidden, so the cue isn't there.
Strong · a meta-analysis
Deciding ahead of time what you'll do
An if-then plan, "when X happens, I'll do Y", had a medium-to-large effect on reaching goals across 94 tests, one of the strongest results in the field.
In Seiza An if-then plan on each profile, shown on the redirect page, and offered by the weekly report when a pattern is clear.
Gollwitzer & Sheeran, 2006, Advances in Experimental Social Psychology ↗
Strong · decades of studies
A specific goal beats "do your best"
Setting a clear, specific goal leads to doing more than a vague intention to try hard.
In Seiza Session goals: a one-line goal at the start and "did you do it?" after. A switch on each profile, off by default.
Modest · a set of lab studies
Writing a plan down quiets an unfinished task
Making a plan for an unfinished task stops it from pulling at your attention while you do something else.
In Seiza Satellites: a thought for later, launched with Alt+P, back when the session ends.
Masicampo & Baumeister, 2011, Journal of Personality and Social Psychology ↗
Strong · a meta-analysis
Tracking your progress helps you get there
Keeping track of progress helps people reach their goals, more so when it's written down. Across 138 studies.
In Seiza The weekly report, and the sky itself.
Modest · a small study, not yet repeated
Habits take a while, and one missed day doesn't undo one
In the best-known study, a new habit took a median of about 66 days, anywhere from 18 to 254, and missing a day now and then didn't set it back.
In Seiza The sky never shows a broken streak, and nothing here promises a habit in 21 days.
Modest · correlational
Being kind to yourself after a slip
Students who forgave themselves for putting off studying put it off less before the next exam.
In Seiza A session that ends early just ends: no scolding, and nothing lost beyond that session.
Wohl, Pychyl & Bennett, 2010, Personality and Individual Differences ↗
Reasoning · not tested
Starting small makes starting easier
The hard part is usually starting, and a pull that comes from a moment of boredom tends to pass once you're back into the work. This is the two-minute rule from Atomic Habits; it hasn't been tested as such.
In Seiza The quiet timer, a short timer you're free to run past, and the redirect page's five minutes.
James Clear, Atomic Habits (2018)
Strong · a large body of work
People keep at what they chose themselves
Motivation from your own reasons lasts; pressure and control from outside wear it down.
In Seiza Seiza only acts on what you set up, never nudges, and every default can be changed.
Ryan & Deci, 2000, self-determination theory, American Psychologist ↗
02 · Considered and left out
Good ideas that don't belong here
- Planning the week at a fresh start, like a Monday or a new month. Seiza isn't a planning app.
- Break timers between sessions. Short breaks help more with tiredness than with getting things done, and they'd mean more settings.
- A separate rule for forgiving slips. Seiza already doesn't punish them.
- Habit goals, like earning a site by going for a run, are parked until they've been thought through. Pairing a treat with a chore does work, but a larger follow-up found a smaller effect than the first study (Milkman, Minson & Volpp, 2014).
03 · What hasn't held up
Popular, but not built on
- Willpower as a fuel tank that runs out ("ego depletion"). A replication across 23 labs found an effect of about zero (Hagger et al., 2016).
- "21 days to form a habit." It doesn't come from research; the study above found anywhere from 18 to 254.
- Deadlines you set yourself beating procrastination, often cited for tools like this one. A replication didn't find it, and the authors have asked for the paper to be retracted (Data Colada). The study of limits you set yourself, above, is the one to go by.
04 · What's kept out
Things Seiza won't do
- Points, badges and leaderboards. Rewards you expect and can count can wear down your own reasons for doing something (Deci, Koestner & Ryan, 1999; still debated). The sky reflects what you did; it isn't a currency to chase.
- Streaks you're afraid to lose. You'll see your current and longest streak, never a broken one.
- Nudges and nagging. Seiza only acts on what you set up.
- The slot-machine pull. Surprises like comets and meteor showers come only from focusing, never from visiting a site, and nothing tells you to chase them.
Each finding was checked against its paper in October 2026. If you know of research that changes any of this, open an issue.