DEV Community: Eusebiu Balan

What 4,000 Habit Trackers Reveal About Self-Improvement

Eusebiu Balan — Sat, 03 Oct 2026 23:36:41 +0000

TL;DR. Across 4,000+ people and 45,000+ habit check-ins, the data tells an uncomfortable story. Most tracked habits are abandoned within days. The habits that last are the low-effort or automatic ones, not the ambitious ones. And 60% of people who sign up to track a habit never log a single check. Measured behaviorally, self-improvement looks far less like steady upward progress and far more like a series of short-lived restarts. That sounds bleak. It is actually freeing, and it tells you exactly what to change.

Here is the most honest sentence I can write about habit tracking, backed by real data instead of a motivational poster: most tracked habits are abandoned within days, the longest-lasting ones are the low-effort or automatic ones, and 60% of people who sign up never log a single habit.

I run Loggd, a habit and life tracker. That gives me a view almost nobody writes about honestly, because it is not flattering: the aggregate behavior of thousands of real people trying to build habits, including all the ones who quit on day one. The self-improvement industry is worth somewhere around $54 to $57 billion a year (Grand View Research; The Business Research Company). The data from inside one tracker suggests most of that money is buying a feeling, not a result.

This is what 4,000+ habit trackers actually reveal.

Do habit trackers actually work?

The honest answer is: yes, but almost never for the reason the app store screenshots imply.

The biggest single predictor of a long streak in our data is not motivation, not premium features, not the perfect reminder time. It is automation. Habits that completed themselves from an outside signal held an average longest streak about nine times higher than habits a human had to remember to check off (more on this below). The tracker helps when it removes friction and shortens the gap between doing the thing and seeing the reward. It does almost nothing when it is just one more box you have to remember to tick.

The cleanest proof of that is the 60% of people who never tick the box at all.

The uncomfortable headline: most habits die in days

Here is the survival curve. We looked at the longest streak each of 5,251 habits ever reached. Not the current streak, the all-time best. The single best run that habit ever had.

Best streak the habit ever reached	Share of all habits
Never checked once	46.4%
Exactly 1 day	28.8%
2 days	6.2%
3 to 6 days	10.5%
7 to 13 days	4.4%
14 to 29 days	2.2%
30 to 65 days	1.0%
66+ days	0.5%

Read that top row again. Almost half of all habits people create never receive a single check-in. The habit is created in a hopeful moment and then never touched.

Stack the next rows and it gets starker. About three out of four habits never make it past a one-day streak. Roughly 90% never reach a single week. And the 66-day mark, the number popularized from Phillippa Lally's 2010 study on how long it takes a behavior to become automatic, is reached by about 0.5% of habits. Half of one percent.

Lally's study found it took a median of 66 days for a new behavior to reach automaticity, with a range from 18 days to 254 days depending on the person and the habit. Our data does not contradict that. It shows that almost nobody gets far enough into the curve to find out where they personally land.

The myth says "do it for 66 days and it sticks." The data says the overwhelming majority of attempts end in the first week, so the 66-day finish line is irrelevant to how habits actually fail. People do not fail at day 66. They fail at day 2.

What people try to build (and what they actually pick)

Before we look at why habits die, it is worth seeing what people reach for. Grouping habit names into rough themes:

Theme	Approximate share of habits
Fitness (gym, exercise, running, steps, stretching)	~27%
Mind and learning (reading, studying, coding, journaling)	~15%
Health and body (water, sleep, vitamins, food, teeth)	~14%
Mindfulness (meditation, gratitude, breathing)	~3%
Digital and focus (less social media, fewer screens)	~3%

Methodology note, said plainly: many of these habit names come from onboarding templates the app suggests, like "Go to the gym" or "Drink 8 glasses of water." So this table reflects what people pick from a list at least as much as what they independently decide to track. The themes are real signal; the exact name counts are not. We will not pretend a templated "Exercise 30 minutes" appearing hundreds of times means hundreds of people independently typed that phrase.

The takeaway survives the caveat: people overwhelmingly try to build fitness and self-discipline habits, the most willpower-intensive category there is. That choice is part of why so many fail. We aim the hardest possible habits at the most fragile possible willpower and act surprised when it does not hold.

Where they fail: the willpower tax

Here is the most useful finding in the whole dataset, and it reframes everything else.

We split habits into two groups: ones a person had to manually check off, and ones that completed automatically by syncing from an external source (in our case, a developer's GitHub commit history, which auto-marks a "code today" habit when real commits show up).

	Manually checked habits	Automatically synced habits
Average longest streak	~2.2 days	~20.5 days
Reached a 7-day streak	6.7%	59.6%

The gap is not subtle. Automatic habits lasted roughly nine times longer on average, and were nearly nine times more likely to survive a full week. Same app, same humans, same notification system. The only thing that changed was whether the habit required an act of willpower to record.

Zoom out across every check-in and the same story repeats. Of the 45,000+ checks logged, about 47% came from automatic syncing rather than a person tapping a button. A tiny minority of habits (the automatic ones) produce nearly half of all the activity, precisely because they do not depend on a human remembering.

This is the willpower tax. Every habit that requires you to remember, decide, and act has a daily failure point built into it. Remove the human step and the habit stops dying. That is not a motivation problem you can solve with a better quote. It is a friction problem you solve with design.

The consistency patterns: when habits slip

Two more patterns, both small and both telling.

Habits slip on the weekend. Check-ins decline almost perfectly from the start of the week to Saturday:

Day	Share of all check-ins
Monday	16.2%
Tuesday	15.7%
Wednesday	15.2%
Thursday	14.9%
Friday	14.5%
Saturday	11.5%
Sunday	12.0%

Monday is the high point of intention. By Saturday, check-ins drop by nearly a third. The "fresh start" energy is real and it is front-loaded into the work week. The honest design response is not to guilt people about weekends, it is to expect the dip and not punish it.

People who also track focus time hold longer streaks. Among users who completed at least one focus session, the average longest habit streak was about 5.2 days, versus about 1.6 days for people who never used the focus timer (based on 400+ focus users, comfortably above our reporting floor).

Be careful with that one. This is correlation, not causation. People who run focus sessions are probably more engaged and more deliberate to begin with, and that underlying trait likely drives both behaviors. Focus tracking is not a magic streak booster. But the association is consistent with the broader theme: people who build a system, more than a single isolated habit, stick around longer.

What it means (the part that is actually freeing)

Put the findings together and the standard self-improvement narrative falls apart:

It is not a discipline problem. 60% never even start, and the ones who do mostly fail in the first two days. That is too fast to be about willpower running out. It is about the system never engaging.
The winning habits are the easy and automatic ones, not the heroic ones. Ambition predicts failure. Friction predicts failure. Low friction predicts survival.
Consistency is fragile and patterned. It dips on weekends, it leans on systems, and it collapses the moment a habit depends on you remembering.

Here is why that is freeing rather than depressing. If self-improvement were purely about willpower, the only fix would be to become a more disciplined person, which nobody knows how to do on demand. But the data says the levers are mostly structural:

Make it automatic where you can. Pick habits that can be triggered or measured by something you already do. The closer a habit gets to "happens whether I think about it or not," the longer it survives.
Start humiliatingly small. A two-minute version of a habit beats a one-hour version that you abandon in three days. The survival curve rewards the trivial.
Stop treating a missed day as failure. This is the big one. A streak that resets to zero on the first miss is a design that manufactures quitting. After one bad weekend, the counter says "0" and people delete the app rather than face it.

That last point is exactly why Loggd's default view is a contribution grid instead of a streak counter. A missed day becomes one lighter square in a year of darker ones, not a reset to zero. The data above is the argument for that design: when 75% of habits never pass a single day, a model that punishes the first miss is a model optimized to make people quit.

Methodology and what this data is not

This is a flagship claim, so it gets a real methodology section, including the parts that weaken it.

Source and scale. Aggregate, anonymized data across 4,000+ Loggd users and 45,000+ habit check-ins, re-run June 2026. All figures are rounded aggregates. No individual data, no user-entered text, no identifying detail.
Selection bias. Everyone here chose to use a habit tracker. That over-represents motivated, self-improvement-minded people and under-represents the general population. If anything, the real-world quit rates are likely worse than what a self-selected tracker audience shows.
Tracked is not the same as done. A missing check-in does not prove the behavior did not happen. Someone may have gone to the gym and not logged it. We are measuring tracking behavior, which is a proxy for, not a perfect record of, actual behavior.
Some quitting is healthy. Abandoning a habit can be a good decision. People drop habits that stopped serving them, or that they replaced with something better. "Quit" is not always "failed."
Seeded templates. Many habit names come from onboarding suggestions, so name-frequency counts reflect menu design as much as personal choice. We report themes, not name-level rankings, for that reason.
Automation sample is small but stark. The automatically synced habits are a small subset (developers syncing GitHub). The 9x effect is large and directionally clear, but it comes from a specific population. Two honest confounds: these are people whose day job already involves coding daily, so part of the consistency is the underlying behavior, not just the automation. And to be clear about how the number is built, it compares the all-time best streak (longest_streak) of every GitHub habit against every manual habit, with no dormant accounts filtered out. When a syncing user goes inactive we pause their rewards and reset their current streak, but their historical best is preserved and still counted here, so this is not a survivorship artifact that quietly drops the quitters. Read the 9x as "removing friction dramatically helps," not as a precise universal multiplier.

None of these caveats overturn the headline. They sharpen it. Even among a motivated, self-selected audience, with tracking-not-doing working in their favor, most habits still die in days. That is not a measurement artifact. That is the reality the motivational version papers over.

Frequently asked questions

Do habit trackers actually work?

They work when they reduce friction and shorten feedback, not when they just add another box to remember. The strongest predictor of a long streak in our data was automation, not features or motivation. The 60% who never log a single check show that the tracker by itself does nothing.

How long do most habits last?

Most do not last. About 46% of habits never get a single check, roughly 75% never pass a one-day streak, and only about 0.5% reach 66 days.

What habits last the longest?

The low-friction and automatic ones. Automatically synced habits averaged a longest streak about nine times higher than manual ones. Easy beats ambitious.

Why do most people fail at self-improvement?

Two reasons, neither of them "laziness." Sixty percent never start (activation failure), and among those who do, the habits that demand daily willpower collapse fast while automatic or trivial ones survive.

Is tracking your habits worth it?

Yes, if you track to learn rather than to perform. The honest picture of what you actually do is the value. A forgiving view keeps you in the data long enough to learn from it; a punishing streak counter just makes people quit.

How was this data collected?

Aggregate, anonymized data across 4,000+ Loggd users and 45,000+ check-ins, re-run June 2026, all rounded, no individuals. The sample is self-selected and many habit names are seeded from templates, both of which are disclosed above.

Written by Eusebiu, who builds Loggd, the habit and life tracker this data comes from.

How Long It Takes to Build a Habit: Data vs Science

Eusebiu Balan — Wed, 30 Sep 2026 23:39:50 +0000

TL;DR. Research says it takes about 66 days on average to make a behavior automatic, not the popular "21 days," with a real range of 18 to 254 days (Lally et al., 2010). But Loggd data on 5,491 real habits shows almost nobody gets that far: only about 0.6% of habits ever reached a 66-day streak, and roughly 46% never reached even a one-day streak. Habit formation takes longer than most people think, and the hard part is not the finish line. It is surviving the first two weeks.

How long does it take to build a habit?

About 66 days on average, not 21. That is the headline finding from the most-cited study on the topic, and the "21 days" number you have heard everywhere is a myth (more on where it came from below).

But the average hides the real story, which is the spread. In the same research, individuals took anywhere from 18 days to 254 days to make a new behavior feel automatic. So the honest answer to "how long does it take to build a habit?" is: somewhere between three weeks and eight months, depending on you and on how hard the habit is.

And then there is what actually happens in the wild. We track habits for thousands of people, and when we look at how far real habits get, the picture is sobering: most never come close to 66 days, because most are abandoned in the first week. The timeline is not the problem. Lasting long enough to reach it is.

The 66-day study, explained (Lally 2010)

The number everyone cites traces back to one paper: Lally, van Jaarsveld, Potts and Wardle (2010), published in the European Journal of Social Psychology. Researchers at University College London asked 96 people to pick a new daily habit (something like "drink a glass of water after breakfast" or "go for a walk before dinner") and tracked them for 12 weeks. Each day, participants recorded whether they did the behavior and how automatic it felt.

By fitting a curve to the rising automaticity scores, the researchers found:

The average time to reach peak automaticity was 66 days.
The range ran from 18 days to 254 days.
Missing a single day did not meaningfully hurt the process. One slip is not a reset.
Some participants never fully automated their habit inside the 12-week window.

That last detail matters. The study's real message is not "it takes 66 days." It is "it varies enormously, and it is slower and more forgiving than you think." The 66 is just an average pulled from a wide, messy distribution.

A 2024 meta-analysis of health-behavior habit formation (PMC11641623) reinforced the same point: the time to form a habit clusters in a broad range rather than landing on any single magic number, and it depends heavily on the behavior and the person. Habit formation is a distribution, not a deadline.

Why "21 days" is a myth

The "21 days to form a habit" rule is one of the most repeated pieces of self-help advice, and it is based on a misreading.

It comes from Dr. Maxwell Maltz, a plastic surgeon who wrote in his 1960 book Psycho-Cybernetics that it took his patients "a minimum of about 21 days" to get used to a change, such as adjusting to a new face after surgery or a phantom limb sensation after an amputation. Read that carefully:

He said "a minimum of about," describing a floor, not a fixed rule.
He was describing adjusting to a change, not building a new behavior from nothing.
It was a clinical observation, not an experiment.

Over the decades, "a minimum of about 21 days" got flattened into "21 days to form a habit," and a casual note became a fake law repeated in thousands of articles. The actual evidence puts the average at roughly three times that long.

Here is the practical danger of the myth: if you believe a habit takes 21 days, you expect to be "done" at three weeks. When it still feels like effort on day 22, you conclude it failed and quit, right around the point where the real work is just beginning.

The number	Where it comes from	What it actually means
21 days	Maltz, Psycho-Cybernetics (1960)	A surgeon's casual note that patients took "a minimum of about 21 days" to adjust to a change. Not a rule.
66 days	Lally et al. (2010), UCL	Average time to automaticity across 96 people. A midpoint, not a guarantee.
18 to 254 days	Lally et al. (2010), the full range	The real spread. This is the honest answer.

What Loggd's real streak data shows (the gap)

Here is where original data beats another retelling of the same two studies. We looked at the all-time best streak of every habit on Loggd and counted how many ever reached each milestone. The sample: 5,491 habits across 2,983 users, re-run in June 2026.

The average longest streak across all habits was 2.8 days. Not 2.8 weeks. Days.

The survival curve, milestone by milestone:

Milestone	Habits that reached it	Share of all habits
At least a 1-day streak	2,973	~54%
7-day streak	457	~8.3%
21-day streak ("the myth")	141	~2.6%
30-day streak	91	~1.7%
66-day streak (automaticity)	31	~0.6%

Read the top row again: about 46% of habits never reached even a one-day streak. People create the habit, fully intend to do it, and then never log a second consecutive day. By the 7-day mark, more than 90% of habits have dropped off. By 66 days, the research's average finish line, only about 1 in 175 habits is still standing.

This is the gap between the science and the reality. The 66-day study describes what happens if you keep going. The data describes how rarely people keep going at all. Both are true, and together they say something more useful than either alone: the timeline was never the hard part. The first two weeks are.

The one thing that beat the curve: removing friction

There is a striking exception in the data, and it points straight at the fix. Some Loggd habits are checked off automatically from an outside data source instead of by hand (for example, a "code every day" habit that completes itself from a GitHub commit log).

Manually checked habits averaged a longest streak of about 2.3 days.
Automatically synced habits averaged about 20 days, roughly nine times longer.

Automated habits are a small slice of the total (under 3% of all habits), so treat this as directional, not a precise universal multiplier, and note the obvious confound: someone syncing a coding habit already codes daily for work, so part of that consistency is the underlying behavior, not just the automation. But the direction is unmistakable. The habits that did not depend on remembering, on willpower, on opening an app, are the ones that survived to the timelines the research talks about. Friction is the thing that kills habits before they can form.

How to actually reach automaticity

If the science says 66 days and the data says almost nobody gets there, the practical question is: how do you become one of the few who does? The data points to a clear answer, and it is the opposite of "try harder."

Plan for two months, not three weeks. Expect a habit to still feel like effort at day 22. That is normal, not failure. Knowing the real number (closer to 66 days) keeps you from quitting at the exact moment most people do.
Start absurdly small. The automated habits won by removing effort. You can copy that manually by shrinking the habit until it is almost impossible to skip: two push-ups, one page, a single glass of water. A tiny habit you actually do beats an ambitious one you abandon in three days.
Anchor it to something you already do. Attach the new behavior to an existing cue ("after I pour my morning coffee, I write one sentence"). A stable cue is what the 66-day research found drives automaticity. No cue, no habit.
Treat a missed day as a single light square, not a reset. The Lally study found that one missed day did not derail the process. The danger is the all-or-nothing reaction: miss once, feel like a failure, quit entirely. A model that resets your streak to zero on the first slip is a model that manufactures quitting.

That last point is exactly why Loggd's default view is a contribution grid instead of a streak counter. A missed day becomes one lighter square in a year of darker ones, not a zero that wipes out your progress. When the data shows that 46% of habits die before a single repeat and only 0.6% reach automaticity, a tracker that punishes the first miss is optimized for exactly the wrong outcome. For the fuller picture of what these patterns look like across thousands of people, see what 4,000 habit trackers reveal.

Methodology and caveats

Because this article makes a data claim, here is what the numbers are and are not:

Source and scale. Aggregate, anonymized data across 5,491 habits and 2,983 Loggd users, re-run June 2026. All figures are rounded aggregates. No individual data, no user-entered text, no identifying detail.
Milestones use all-time best streak. "Reached 66 days" means a habit's longest_streak value hit 66 at some point, not that it is still active. This is the most generous possible reading, and even so only 0.6% qualify.
Selection bias. Everyone here chose to use a habit tracker, which over-represents motivated, self-improvement-minded people. Real-world quit rates are probably worse, not better, than what a self-selected tracker audience shows.
Tracked is not the same as done. A missing check-in does not prove the behavior did not happen. We are measuring tracking behavior, a proxy for actual behavior.
The automation effect comes from a small group. The 9x figure compares a small subset (developers syncing GitHub) against manual habits, with the day-job confound noted above. Read it as "removing friction helps a lot," not as an exact law.

None of these caveats overturn the headline. They sharpen it. Even among motivated people, with tracking-not-doing working in their favor, the overwhelming majority of habits never get within sight of 66 days.

Frequently asked questions

How long does it take to build a habit?

On average about 66 days (Lally et al., 2010), with a real range of 18 to 254 days. The "21 days" figure is a myth. In our data, only about 0.6% of habits ever reached a 66-day streak, because most are abandoned in week one.

Is the 21-day rule real?

No. It is a misreading of Maxwell Maltz, who in 1960 observed patients taking "a minimum of about 21 days" to adjust to a change. That floor became a fake rule. The research average is closer to 66 days.

What is the 66-day study?

Lally, van Jaarsveld, Potts and Wardle (2010), a UCL study of 96 people that found an average of 66 days to reach automaticity, with a range of 18 to 254 days and one missed day having little effect.

How many days to break a bad habit?

There is no clean number; the same 18-to-254-day range is the best guide. Breaking is harder than building, and replacing the routine with a small alternative works better than white-knuckling it.

Why do habits take so long?

Because a habit is a brain shortcut that only forms through repeated reps under a stable cue, and that rewiring is gradual. It also breaks easily early, which is why most habits die in the first two weeks.

Do habit trackers speed it up?

They can, by adding a feedback loop and a small reward and by making your real pattern visible. The biggest lever in our data was reducing friction: automated habits lasted roughly nine times longer than manual ones.

Written by Eusebiu, who builds Loggd, the habit and life tracker this data comes from.

The 50 Most-Tracked Habits on Loggd (2026 Data)

Eusebiu Balan — Sun, 27 Sep 2026 00:07:22 +0000

TL;DR. Based on 6,700+ habits tracked by 3,600+ Loggd users in 2026, the most-tracked habits are going to the gym, exercising for 30 minutes, walking 10,000 steps, and drinking more water, followed by reading, journaling, and meditation. But raw popularity is misleading. Several of the top entries are habits the app suggests during onboarding, and the gap between the habits people start and the ones they keep is large: most tracked habits never reach even a 2-day streak. The habits that actually survive are the ones that log themselves. Here is the full ranked list, plus the honest read of what it means.

Around 280 people on Loggd track "go to the gym" in some form. It is the single most-tracked habit on the platform. And it has one of the shortest average streaks on the entire list.

That contradiction is the real story here. This is not a prescriptive "here are good habits to track" listicle, every habit app publishes one of those. This is a descriptive look at what thousands of real people actually chose to track, which ones the app nudged them toward, and the large gap between starting a habit and keeping it.

What are the most-tracked habits right now?

Here are the top 10 most-tracked habits on Loggd in 2026, by number of distinct users. Counts are rounded. The last column marks whether the habit is one Loggd offers as a one-tap suggestion during onboarding (more on why that matters below).

Rank	Habit	Users (approx.)	In onboarding picker?
1	Go to the gym	280	Yes
2	Exercise 30 minutes	215	Yes
3	Gym	155	No (typed)
4	10,000 steps	150	Yes
5	GitHub activity	135	Yes (auto-syncs)
6	Drink 8 glasses of water	115	Yes
7	Workout	95	No (typed)
8	Morning water	90	Yes
9	Morning stretch	85	Yes
10	Read 30 minutes	80	Yes

Two things stand out immediately. First, fitness and hydration dominate. If you group the obvious synonyms ("go to the gym," "gym," "workout," "exercise," "exercise 30 minutes"), exercise is by far the most-tracked theme on the platform, ahead of everything else combined. Water is second. Reading is third.

Second, eight of the top ten are one-tap suggestions from Loggd's onboarding picker. Only "gym" and "workout" (the bare, user-typed variants) are not. That is the single most important thing to understand before you read any "most popular habits" list, including this one: popularity is heavily shaped by what the app puts on the menu. We come back to this below.

The one genuine outlier in the top 10 is GitHub activity (5th). It is on the menu, but it is not a manual habit at all, it auto-logs from a developer's commit activity. Hold that thought, because it turns out to be the most important habit on this entire list.

The full top-40 list

Past the top 10, the list broadens into the habits you would expect: sleep, vitamins, journaling, reading, coding, and the small daily anchors. Counts are rounded to the nearest five; every entry shown has at least 50 users behind it except where noted.

Rank	Habit	Users (approx.)
1	Go to the gym	280
2	Exercise 30 minutes	215
3	Gym	155
4	10,000 steps	150
5	GitHub activity	135
6	Drink 8 glasses of water	115
7	Workout	95
8	Morning water	90
9	Morning stretch	85
10	Read 30 minutes	80
11	Exercise	80
12	Run	80
13	Do pushups	70
14	Sleep 8 hours	65
15	Take vitamins	60
16	Journal	55
17	Read	55
18	Meditate	50
19	Read 10 pages	50
20	Post on threads	50
21	Brush teeth	45
22	Practice coding	40
23	No alcohol	40
24	No junk food	35
25	Reading	35
26	In bed by 10pm	35
27	Study	30
28	Drink water	30
29	Meditation	30
30	10k steps	25
31	Deep work session	25
32	Plan tomorrow	20
33	No social media	20
34	Practice language	20
35	Skincare routine	20
36	Eat fruit	20
37	Sport	20
38	Morning routine	20
39	Take the stairs	15
40	Make bed	15

A few things to read out of the long tail:

The synonyms tell you what people mean, not what they typed. "Gym," "go to the gym," "workout," "exercise," and "sport" are the same intention typed five ways. The category, not the exact label, is the real signal.
The bottom of the list is where it gets interesting. Below rank 40, the counts drop fast and the habits get genuinely personal: take creatine, track expenses, skincare, language practice, take the stairs. The long tail is enormously diverse, which is itself a finding. There is no "correct" set of habits. People track their own lives.

How many of these are app suggestions vs. choices people made?

This is the part most "popular habits" articles leave out, and it is the part that changes how you should read the table.

Most of the top entries are habits Loggd offers as a one-tap suggestion during onboarding. "Go to the gym," "exercise 30 minutes," "10,000 steps," "drink 8 glasses of water," "morning water," "morning stretch," "read 30 minutes," and others all sit in the starter picker a new user sees on day one. When someone is handed a tidy menu of starter habits, a lot of them tap a few without much commitment, then never check them again.

So the popularity ranking reflects menu design as much as personal motivation. A templated "Exercise 30 minutes" appearing 200+ times does not mean 200 people independently decided that was their habit. It partly means 200 people tapped a suggestion during setup. The clearest tell is in the synonyms: "go to the gym" (a picker option, ~280 users) towers over "gym" and "workout" (the typed-from-scratch versions, ~155 and ~95), even though they mean the same thing. The gap is the menu effect.

We are flagging this rather than hiding it, for one reason: a popularity stat that pretends every entry was a deliberate, organic choice is misleading, and a misleading stat is not worth citing. The themes are real signal (people genuinely want to exercise, drink more water, read more). The exact name-level rankings are softer than they look, because the app's own starter menu is sitting on the scale.

The honest version of the headline: the most suggested-and-accepted habits are gym, exercise, steps, and water. Whether people keep them is a separate question, and the answer is sobering.

Which popular habits do people actually keep?

Here is the gap. The most-tracked habits are not the most-kept habits. In fact, they are often the opposite.

Across all 6,700+ habits on Loggd:

About 45% never reach even a 1-day streak. They get created, sometimes from a suggestion, and never checked once.
Most of the rest fade within a day or two. The median habit barely registers a streak at all.
Only a small fraction ever reach a 7-day streak.

And when you sort the popular habits by how long people actually keep them, the most-wanted habits do worst:

Habit	Avg. longest streak (days)	Note
GitHub activity	~21	Auto-synced
Post on threads	~16	Public, low-friction
Take vitamins	~4.5	Tiny, anchored
Workout	~3.4	Manual
Morning water	~3.2	Tiny, anchored
Brush teeth	~3.0	Already automatic
Go to the gym	~1.5	High friction
10,000 steps	~1.1	High friction
Exercise 30 minutes	~0.9	High friction

Two patterns jump out.

First, the two longest-surviving habits both log themselves. GitHub activity (auto-synced from commits) and posting on a public feed have streaks roughly 5 to 15 times longer than any manually-checked habit. The lesson is not about willpower. It is about friction. A habit you do not have to remember to log is a habit that survives. We wrote about this directly in what 4,000 habit trackers reveal: the single biggest driver of a long streak is whether the habit auto-completes.

Second, among manual habits, small beats big. "Take vitamins," "brush teeth," and "morning water" outlast "go to the gym" and "30-minute workout" by a wide margin. The big effortful habits are the ones people most want and abandon fastest. The tiny anchored ones quietly survive.

What this means if you are choosing habits to track

If you came here for a list of habits to copy, here is the more useful takeaway: how you track matters more than what you track.

Start with 1 to 3 habits, not ten. The more habits people start at once, the lower the follow-through. A short list you actually check beats a long list you abandon.
Pick small and anchored over big and effortful. "Drink a glass of water when I wake up" survives. "Go to the gym" does not, at least not as a checkbox. Shrink the habit until it is almost too easy.
Reduce logging friction. The habits that last are the ones that log themselves or take one tap. If checking the box is a chore, the habit dies, regardless of how motivated you were on day one.
Do not let a broken streak end the habit. Most people quit the first time they miss a day. That is exactly why Loggd's default view is a forgiving contribution grid instead of a streak counter that resets to zero. A missed day should be a lighter square in a good year, not a reason to give up.

The most-tracked habits on Loggd are the ones people aspire to. The most-kept habits are the ones people made easy. If you want to be in the second group, choose accordingly.

Methodology

These figures come from aggregate, anonymized data across 3,600+ Loggd users and 6,700+ habits, re-run in June 2026.

What is counted. Habit names were normalized (lowercased, trimmed) and ranked by distinct users, so one power-user cannot skew a name. All published counts are rounded to the nearest five.
Minimum sample. No named habit count is published with fewer than 50 users behind it. Streak averages shown are for habits with a large enough sample to be stable.
Onboarding picker effect. Most of the top habit names are one-tap suggestions in Loggd's onboarding picker ("go to the gym," "drink 8 glasses of water," "morning water," "read 30 minutes," and so on), so name-frequency counts reflect menu design as much as personal choice. We report this openly because it changes how the ranking should be read. The themes are real; the exact name-level ordering is partly an artifact of the menu.
Selection bias. The sample is self-selected: people who chose to use a habit tracker. It over-represents the motivated and under-represents the general population.
Privacy. No individual user data, no user-entered free text beyond generic habit labels (which are generic by nature, "gym," "read"), and no usernames or demographics.

On the "which survive" framing: habit automaticity averages around 66 days per Lally et al. (2010), with a wide range (18 to 254 days), and a 2024 meta-analysis found health-habit formation can vary even more widely. So "survives" here means "kept checking," not "became automatic." Most habits on Loggd fade long before either.

Frequently asked questions

What is the most popular habit to track?

Gym and exercise variants dominate, followed by 10,000 steps and drinking water, then reading, journaling, and meditation. Note that several of these are onboarding suggestions, so popularity reflects the menu as much as personal choice.

How many habits should I track at once?

One to three. The more you start at once, the lower the follow-through. Get a couple sticking before adding more.

Are app-suggested habits worse than ones you pick yourself?

Mixed. Suggested habits inflate the popularity numbers because many get added and ignored. But the bigger predictor of survival is logging friction, not who chose the habit.

What habits are easiest to keep?

The ones that log themselves (auto-synced habits have by far the longest streaks) and, among manual habits, small anchored ones like a vitamin or morning water.

What habits do people quit fastest?

High-friction physical habits: 30-minute workouts, 10,000 steps, the gym, running. The most-wanted habits are also the fastest abandoned.

Is tracking habits actually effective?

It helps as a feedback loop, but it is not magic. Most started habits still fade fast, and real automaticity averages around 66 days.

How was this data collected?

Aggregate, anonymized data across 3,600+ users and 6,700+ habits, re-run June 2026, all rounded, 50-user minimum per named stat, no individuals. Most top habit names are one-tap suggestions from Loggd's onboarding picker, which is disclosed above.

Written by Eusebiu, who builds Loggd, the habit and life tracker this data comes from.

Habits People Quit Fastest (7,800 Habits of Data)

Eusebiu Balan — Thu, 24 Sep 2026 00:04:14 +0000

TL;DR. About 45% of tracked habits are never checked off once, and another 23% are checked exactly one time and never again. That is the pattern across 7,800+ real habits and 4,200+ people, and it looks the same in every habit tracker; this one just publishes it. Roughly two in three habits die at or before the second check-in. The habits abandoned fastest are the high-effort physical ones: running, morning stretch, pushups, 30-minute workouts, 10,000 steps, and reading 10 pages, all averaging streaks of a day or less. The popular story is that people quit at day three, or that a broken streak kills a habit. The data says something less flattering and more useful: most habits never get far enough to break a streak. They are created, checked once, and quietly abandoned.

If you have ever set up a habit tracker with real intent and then found it untouched two weeks later, you are looking at the single most common outcome in the data. Not the exception. The default.

I run Loggd, a habit and task tracker, so I can see the aggregate shape of this across thousands of people. Here is what actually happens to a habit after someone creates it.

One thing to say up front, because it matters for how you read every number below. This is not a Loggd problem, it is the shape of habit formation itself. Every habit tracker on the market has numbers like these, and the published research points the same way: the Lally study at University College London found automaticity takes a median of 66 days, which almost nobody reaches in any system. The difference is that most apps only ever show you the success stories in their marketing. I would rather publish the real curve, because knowing where habits actually break is the only way to design around it, both for you and for the app.

Which habits do people quit fastest?

Sorting habits by their average longest streak, and only counting habit names that at least 50 different people track, the bottom of the list looks like this.

Habit	People tracking it	Avg. longest streak (days)
Run	90	0.7
Morning stretch	100	0.7
Do pushups	85	0.9
Exercise 30 minutes	245	0.9
10,000 steps	180	1.0
Read 10 pages	50	1.1

An average longest streak of 0.7 days means that across everyone tracking "run," the typical best-ever run of consecutive days did not reach a single day. Most of those habits were never checked at all.

Two things stand out. First, every entry is a physical or scheduling habit that demands a real block of time or a real change to your evening. Second, "read 10 pages" is on this list despite being the kind of habit that usually gets recommended as an easy win. Ten pages is not small when you are starting from zero pages.

We covered the flip side of this list in the 50 most-tracked habits on Loggd: the habits with the longest streaks are the ones that log themselves, like GitHub activity syncing automatically. The friction of logging predicts survival better than the ambition behind the habit.

But the habit names are the least interesting part of this data, because sorting by name hides the real finding.

Most habits are never checked twice

Instead of streaks, count raw check-ins. How many times was each habit ever marked complete?

Total check-ins	Share of all habits
0	~45%
Exactly 1	~23%
Exactly 2	~7%
3 to 6	~10%
7 to 30	~10%
31 or more	~5%

Read the top two rows together. About 68% of habits accumulate one check-in or fewer. Nearly half never get a single one.

Put differently: of the habits that do get checked off at least once, about 41% never receive a second check. The person set it up, did the thing once, felt the small hit of satisfaction, and never came back.

This reframes the whole conversation about quitting. The popular narrative, the one every motivation blog runs on, is that people build momentum for a few days and then lose it. That story assumes a stretch of consecutive days that eventually breaks. For two thirds of habits, that stretch never exists. There is no streak to break.

The wall is not day three. The wall is check two.

How long does a habit actually live?

For habits that got at least one check-in, measure the days between the day the habit was created and its final check-in. That is the habit's lifespan.

Lifespan	Share of habits that were checked at least once
Last check on the day it was created	~33%
1 to 2 days	~13%
3 to 6 days	~10%
7 to 29 days	~18%
30+ days	~26%

A third of all habits that ever get checked are checked only on the day they were created and never again. Setup and first check-in are the same event, and it is also the last event.

One honest caveat on that bottom row: the 30+ day group includes habits that are still running right now, so it mixes long survivors with people who are simply mid-habit. It is not a clean "these lasted a month and then stopped" figure. The top rows are unambiguous, though, because a habit last checked on its creation day two months ago is genuinely finished.

Set against the streak data from how long it takes to build a habit, where reaching the 66-day mark that the Lally research associates with automaticity is a sub-1% event, the picture is consistent. Almost nobody gets near the point where a behavior becomes automatic, because almost nobody gets past week one.

Are app-suggested habits abandoned more often?

Yes, and it is worth measuring rather than assuming, because almost every tracker offers a starter menu.

Loggd, like almost every tracker, shows new users a list of suggested habits: go to the gym, drink 8 glasses of water, morning stretch, read 30 minutes. Tapping one takes a second and carries no real commitment, so some of those habits were never genuinely chosen. That should show up in the data, and it does:

Habits picked from a suggestion menu: never checked 52% of the time.
Habits people typed themselves: never checked 43% of the time.

A nine point gap. Choosing your own habit measurably beats picking one off a list, which is a useful finding if you are setting up any tracker: type your own, in your own words, even if the menu has something close.

What the gap does not do is explain the pattern. Habits that someone thought about, typed out, and named themselves are still abandoned before the first check-in more than four times in ten. Remove suggestion menus entirely and the headline finding barely moves. People create habits they do not repeat, whether or not an app proposed them.

Why the second check is the real wall

Three things separate a habit that gets a second check from one that does not, and none of them are motivation.

The first check is free, the second one is not. Creating a habit and checking it once happens in a burst of intent, usually in one sitting. The second check requires a completely different thing: remembering, on a different day, in a different mood, with the novelty gone. Most habit advice optimizes for the burst. The burst is not the problem.

Size is decided at creation, and it is usually wrong. "Exercise 30 minutes" at 245 people tracked and a 0.9-day average streak is the clearest example in the dataset. Thirty minutes is a decision made by an optimistic version of you who is not the person who has to show up tomorrow. The habits that survive tend to be small enough that a bad day does not disqualify them.

Friction compounds daily, ambition does not. Auto-synced habits on Loggd, like GitHub activity, run streaks many times longer than manual ones, as we broke down in what 4,000 habit trackers reveal. Those users are not more disciplined. Their habit costs nothing to log. Every unit of friction you remove is paid back every single day; every unit of ambition you add is charged every single day.

There is also a structural reason streak-based tracking makes this worse. If your tracker shows a streak counter, a missed day resets it to zero, and the visible zero reads as failure. That is the abstinence-violation effect in software form. It is exactly why Loggd's default view is a forgiving contribution grid rather than a streak counter: a missed day should be one lighter square in a year of squares, not a reset.

How to pick habits you will not quit

Short version, based on what the surviving habits have in common:

Make the second check trivial, not the first. Ask what version of this you could do tomorrow on your worst day, then make that the habit.
Anchor it to something you already do daily. After I pour my coffee, after I brush my teeth, after I close my laptop. An existing routine carries the remembering for you.
Track one to three habits, not ten. A long list is the fastest way to check none of them.
Remove logging friction. One tap, a widget, or an integration that checks itself. If logging is a chore, the habit dies regardless of how much you wanted it.
Treat a missed day as a lighter square. Never miss twice is a rule you can keep. A perfect streak is not.

The full playbook, with the specific fixes for each failure point, is here: how to stop quitting habits after 3 days.

FAQ

Why do I quit habits so fast?
Because the habit was almost certainly too big to repeat the next day. In this dataset, about 45% of tracked habits are never checked once and another 23% are checked exactly once, so roughly two in three die at or before the second check-in. This is the normal outcome, not a character flaw.

What habits do people abandon fastest?
High-friction physical habits: running, morning stretch, pushups, 30-minute workouts, 10,000 steps, and reading 10 pages, all averaging streaks of around a day or less.

Is it normal to quit a habit after 3 days?
Reaching day three puts you ahead of most tracked habits. Only about 26% ever get three or more check-ins.

Does habit tracking make quitting more likely?
Streak-based tracking can, because one missed day resets a visible counter to zero and triggers an all-or-nothing response. A forgiving grid view avoids that failure state.

Is the 21-day rule real?
No. It traces back to a misreading of Maxwell Maltz's 1960 observations about patients adjusting to surgery, not to habit research. The Lally 2010 study at UCL found a median closer to 66 days, with a range from 18 to 254 days depending on the behavior.

How do I stop quitting habits?
Optimize for the second check-in: shrink the habit, anchor it to an existing routine, and use a tracker that does not punish a missed day.

Methodology

Figures come from aggregate, anonymized data across 7,800+ habits and 4,200+ Loggd users, re-run in August 2026.

What is counted. Completed check-ins per habit, and the number of days between a habit's creation date and its final completed check-in. Habit names were lowercased and trimmed before grouping.
Minimum sample. No named-habit statistic is published with fewer than 50 people behind it, counted as distinct users. All figures are rounded.
Onboarding picker effect. Many habit names come from Loggd's onboarding suggestion menu. We measured that effect directly and report it above rather than burying it.
Survivorship. The 30+ day lifespan bucket includes habits that are still active, so it is not a pure "quit after a month" figure.
Selection bias. This sample is people who chose to install a habit tracker. It over-represents the motivated.
Privacy. No individual user data, no user-entered free text beyond generic habit labels, no demographics.

Written by Eusebiu Balan, a developer of nine years, now building Loggd full time. I publish the numbers out of the app as I find them, including the unflattering ones.

[Boost]

Eusebiu Balan — Wed, 23 Sep 2026 23:50:39 +0000

Top comments reveal modern coding lost its magic

James Anderson

Sep 23

Something About Coding Stopped Feeling Good — and It Took Me a While to Figure Out What

#discuss #career #ai #watercooler

6 min read

Never Miss Twice: What 2,066 Habits Say About the Second Day

Eusebiu Balan — Sun, 13 Sep 2026 00:10:57 +0000

TL;DR. I measured the "never miss twice" rule against 2,066 real habits from 1,006 people. Once a habit has one missed day behind it, 80.9% of those gaps still end with a check. At two missed days it is 71.1%, a drop of about 10 percentage points, and that is the largest single-day drop anywhere in the curve. So the rule is aimed at something real. But there is no cliff at two. By seven missed days the figure is 42.2%, by fourteen it is 26.7%, and every day in between costs another 5 to 8 points. The honest version of the rule is not "two is fatal". It is that the return trip gets more expensive every single day, which makes the cheapest day to come back always today.

The rule is four words long. Never miss twice. One missed day is an accident, two in a row is the start of a different pattern, so whatever else happens, do not lose the second day.

It is James Clear's line from Atomic Habits, and it travels well because it is short and because it gives you something to do the morning after a bad day, which is more than most habit advice manages. What I had never seen was anyone check it against a large pile of real check-ins. I build Loggd, so I have one.

The question I put to the data

Not "how long do streaks last", but a narrower one, more useful if you are the person staring at yesterday's empty square:

Given that a habit has already gone k days without a check, how often does it ever get checked again?

To answer it I took every stretch of consecutive missed days across 2,066 manual habits belonging to 1,006 people, and asked how each stretch ended. Either the person came back and checked, or the stretch was still open on the snapshot date. The percentage below is, of every stretch that reached k missed days, the share that eventually ended with a check.

Days missed so far	Eventually checked that habit again
1	80.9%
2	71.1%
3	63.3%
4	58.3%
5	52.5%
6	47.3%
7	42.2%
10	34.3%
14	26.7%

The ends of that curve are not thin. The one-day row rests on 10,808 separate gaps, 8,742 of which ended with a return. The seven-day row is 3,144 gaps and 1,326 returns. The fourteen-day row is 2,312 gaps and 618 returns.

The finding is the slope, not any single number

Read the table as differences rather than levels and the shape gets clearer.

From one missed day to two, the return rate falls by about 10 percentage points. That is the largest single-day drop anywhere in the curve. Every subsequent day costs roughly 5 to 8 points.

So the second day is genuinely the worst single day to lose. If you were going to hang a rule on one day, that is the right day.

But look at what happens after it. Nothing dramatic. Day three costs you, day four costs you, day six costs you. The line keeps sliding at a fairly steady rate all the way out to two weeks. There is no floor where the habit is officially dead and no ledge to fall off. The curve describes a slope with no bottom in sight.

What this supports, and what it does not

Supported: the second day is the sharpest single decision point, so a rule that puts all its weight there is putting it somewhere defensible. If you remember one thing about missed days, "do not lose the day after" is the right thing to keep.

Not supported: the cliff. The way "never miss twice" gets repeated implies that one miss is fine and two is catastrophic, which turns the rule into another version of the all-or-nothing thinking it was supposed to replace. If you already missed twice, the rule as commonly stated has nothing left to say to you. You failed its only test on day two, out of a possible fourteen where the numbers say returning is still ordinary. Look at that fourteen-day row again: 26.7%, which is 618 real returns from a two-week hole.

So the version I would actually give someone is this: the return trip gets harder every single day, so the cheapest day to come back is always today. It says the same thing about day two that Clear's version says. It also still works on day ten, which his does not.

What "coming back" should look like

The mistake I see most often, and have made myself, is treating the return as a debt.

You missed Tuesday, so Wednesday has to be a double session: two runs, forty pages instead of twenty, the workout you skipped plus the one you owe. It feels like accountability. What it does is price the return higher than the original habit, on a day you already proved you had no spare capacity, which is why so many of those Wednesdays quietly become Thursdays.

The data has no opinion about the size of the check, only whether one happened. A two-minute version and a full session are the same row in the table.

So make the return the smallest honest version of the thing. The run becomes a walk to the end of the road. Twenty pages becomes one page. The gym session becomes the bag by the door and ten minutes of something. Then mark it and let the day be completely ordinary: no ceremony, no restart post, no recommitment. The point of the small version is that it costs almost nothing, and the whole finding here is about cost.

If you want the fuller treatment of what to change so the miss happens less often in the first place, that is how to stop quitting habits. This post is only about the days after.

Why the record you keep changes the odds

There is a second reason the make-up session is tempting, and it is usually sitting on your screen. A streak counter that shows a month of work and then shows zero has told you the month was worth nothing. That is not a neutral display, it is an argument, and the argument it makes the morning after a miss is that the run is already ruined. Which is the exact story that turns one missed day into two.

A grid makes a different argument. One lighter square in a wall of darker ones is a proportionate account of what happened: you missed a day in a month of days. I have written the full case for that design in contribution grid versus streaks, and the curve above is the strongest evidence for it I have found. If day two is where most people are lost, the display you look at on day one should not be the thing arguing for giving up.

Two of our free tools sit on either side of that line. The habit streak calculator is deliberately streak-shaped: you click the days you did the thing on a calendar and it shows your current run, your longest ever run, and a consistency percentage (days completed divided by days since your start date), all in your browser. The consistency figure is the interesting one, because it keeps counting when the streak resets to zero. The 100 day tracker takes the opposite approach: days made and current run are separate numbers, so a skipped day restarts nothing.

Decide your miss policy before you need it

The version of you that misses a Thursday in November is tired, slightly embarrassed, and unusually receptive to the idea that the whole thing is already over. That is not the person who should be setting policy. So write it down now, while nothing is at stake. Mine is three lines:

A missed day is a missed day. It gets marked honestly and nothing else happens.
The next day is the normal version of the habit, not a bigger one.
If I miss twice, the third day is the smallest version I can think of, and I still do not restart the count.

Rule three is the one the data added. The common phrasing of "never miss twice" leaves you with no move once you have already missed twice, and the numbers say that is precisely when a move is still worth making.

What this data is not

Four things to know before citing any of the above.

This is correlation, not an experiment. Nobody was assigned a missed day. Habits that were already fragile or badly sized are over-represented among the long gaps, so part of what the falling curve measures is which habits were shaky to begin with, not what the missing did to them.

Survivorship. Only habits checked on at least two separate days are in this sample. Habits that were created and never really started are excluded entirely, and that is a much larger group than this one. It is covered in habits people quit fastest, which is where to go for the real abandonment picture.

Censoring. A stretch still open on the snapshot date counts as "not yet returned". That is the correct survival treatment, but it means gaps that started shortly before the snapshot are judged early.

Day attribution is UTC. A late-night check can land on the neighbouring day for some users, which blurs individual gap lengths at the edges.

Methodology

Figures come from a production snapshot dated 2026-08-05, covering 2,066 manual habits with at least two distinct check days, across 1,006 distinct users, with the demo account excluded. Integration-fed habits (GitHub, Threads) were excluded entirely, because they auto-check and would erase every gap they contain. Every stretch of consecutive missed days was counted; a stretch either ended because the person checked in again, or was still open on the snapshot date. Each percentage is the share of stretches reaching that length that eventually ended with a check. No individual user data, no free text, and no habit-level detail is published. The broader dataset these habits sit inside is described in what 4,000 habit trackers reveal.

The one sentence version

Never miss twice is good advice pointed at the right day and wrapped in the wrong story. There is no cliff, only a slope that charges you a few points of your own return odds every day you stay away. Which means the question in front of you is never "is this ruined". It is only ever "is today cheaper than tomorrow". It is. It always is.

Written by Eusebiu, a developer of nine years, now building Loggd full time, so the missed days in my own grid are not hypothetical either.

Why To-Do Lists Fail (21,655 Tasks of Data)

Eusebiu Balan — Sun, 06 Sep 2026 06:50:46 +0000

TL;DR. I counted every task in Loggd. Across 21,655 one-off tasks from 1,375 people, 7,774 were completed, which is 35.9%. The shape matters more than the rate: 40.8% of completed tasks were finished the same day they were created, the median completed task took 1 day, and the ones that did not move quickly now sit a median of 56 days past their own planned date. A to-do list is not a reservoir that drains slowly. It is a queue that either moves today or turns into archaeology.

I build a habit and task tracker, which means I can do the thing most productivity writing cannot. I can open the database and count.

So I did, in August 2026. The raw population is 51,201 tasks across 1,375 users, with my demo account excluded.

Then I threw out more than half of it, and you should know why before you trust a single number below.

The 29,546 tasks I threw out

57.7% of those tasks, 29,546 of them, are not decisions anybody made. They are auto-generated instances of recurring tasks: the "take the bins out" that reappears every Tuesday whether you want it or not.

Recurring instances behave nothing like the things people type into a list at 11pm. The app generates them, not a person having a thought. Counting them would either flatter the completion rate or wreck it, depending on how many instances got spawned for a routine somebody quit in March. Either way the number stops meaning what the headline claims.

So everything from here on is one-off tasks only: 21,655 tasks across 1,375 users. Things somebody deliberately wrote down once.

The uncomfortable number

7,774 of those 21,655 tasks were completed. That is 35.9%.

Most of what people write down never gets ticked off.

The obvious objection is that this is an accounting artefact: plenty of tasks are simply not due yet, and any snapshot catches a pile of healthy pending work. It does not hold here. 13,880 one-off tasks are still open, and this is what they are:

Status of the 13,880 open tasks	Share
Planned day has already passed	94.7%
Planned for a future date	0.7%
Never had a planned day at all	4.5%

Fewer than one in a hundred open tasks is legitimately pending. Recompute the completion rate over settled work only, meaning completed plus overdue plus never-scheduled, and it barely moves: 36.1%.

There is no escape hatch in a "cancelled" bucket either. Exactly 1 task in the entire snapshot is cancelled. Tasks do not get retired, they just sit there.

How old the rot is

This is where it stopped being an abstract percentage for me.

Among the overdue tasks, the median is 56 days past its planned day and the 90th percentile is 132 days. 826 distinct users are carrying at least one.

Two months past a date the person picked themselves. The top tenth is more than four months gone: things written in spring, still nominally scheduled, still producing a small pulse of guilt every time the app opens.

I have several of those. One of them has moved house with me.

The actual finding: tasks are done immediately or never

Here is the part that changed how I think about lists.

Of the 7,774 completed tasks:

40.8% were completed the same day they were created.
The median time from creation to completion is 1 day.
The 90th percentile is 11 days.

Read those three lines next to the 56-day median rot and the shape falls out. There is no slow, steady drain where old tasks get picked off over weeks. Nine in ten completed tasks were done inside 11 days, and 40.8% before the day was over. Anything still open past that window is not in progress. It is archaeology.

That reframes what a to-do list actually is. It is not storage. It is a queue with a very short effective memory, and writing something into it does almost nothing to raise the odds that it happens next month. If it does not move now, or nearly now, the honest forecast is that it never will.

Even the tasks people finished missed their date

One more number, and it is the one that makes me most sympathetic to everyone here.

Take only completed tasks that had been given a planned day: 7,506 of them, across 853 users. How many landed on the day they were planned for?

55.1%. Another 38.7% were completed late and 6.2% early.

So even inside the population of pure successes, nearly half missed their own date. That is not an effort failure, it is an estimation failure, and it is consistent enough that Kahneman and Tversky named it: the planning fallacy, our habit of predicting how long something takes from the version of the story where nothing goes wrong.

The date you assign a task is not a plan, it is a wish with a timestamp. Adding precision to your scheduling is mostly wasted effort. Reducing the number of things you schedule is not.

The aggregate is not you

Now the honest correction, because "35.9% of tasks get done" is the kind of stat that gets screenshotted without its context.

That aggregate is dragged down by one pattern: somebody signs up, dumps forty things they have been carrying around in their head, feels enormously better, and never comes back. Those forty tasks sit in the denominator forever.

Filter to people who used the thing more than once. Among the 340 users who created at least 5 one-off tasks, the median personal completion rate is 54.5%.

That is a very different picture: the typical engaged person finishes a bit more than half of what they write down, which sounds about right for a human with a job.

The tail is real too. 45.3% of those users finish under half of what they write down. So the aggregate is pessimistic, and the individual reality is still that a large minority of people are running a permanent deficit against their own list.

What I would actually change

All of this points at the same four moves, and none of them are "try harder".

1. Cap the day, then let the cap do the deciding. Since almost everything that gets done gets done within a day or so of being written, the useful question is not "what needs doing" but "what am I doing today, given that today is roughly all I get". A fixed small number is the crudest and most effective version of that. The 1-3-5 task planner makes the constraint concrete: one big task, three medium, five small, an optional time estimate per task, and a prompt to carry over yesterday's leftovers. Nine slots, and when they are full they are full. I covered the method itself in the 1-3-5 rule guide.

2. Decide the first action before the day starts. A task that reads "taxes" has no entry point, so it loses to every task that does. The 55.1% on-time figure suggests the weak link is specification rather than willpower. Pick the one thing that must move and name its first physical step. Eat the frog is the least complicated version of that ritual.

3. Let the backlog go, deliberately. Given the 56-day median, a task that is two months overdue is not waiting. It is dead and unburied. Once a quarter, move the whole overdue pile somewhere that is not a to-do list. A brain dump suits this: write everything out without filtering, optionally on a short timer, then tag each line as a task, idea, worry, or reminder. Most of it turns out to be worries and ideas wearing a task costume. The survivors go back on the list with a date. The rest get deleted, which is a decision, not a defeat.

If you would rather sort than dump, the Eisenhower matrix does the same triage across four urgency and importance quadrants, and lets you drag things between them as you change your mind.

4. Plan fewer things rather than better things. Every real improvement I have made to my own list was a subtraction. How to plan your day covers the ritual side, but the arithmetic is blunt: if 35.9% of written tasks get done, writing twice as many does not double your output. It doubles your overdue queue.

You may have seen the Zeigarnik effect cited around this topic, the idea that unfinished tasks stay active in memory and nag at you. It matches how an overdue list feels, but its replication record is mixed rather than settled, so treat it as a decent metaphor for why a stale backlog feels heavy, not as a mechanism you can bank on.

Methodology, and what this data cannot see

The figures come from a production snapshot queried on 2026-08-10, covering data through 2026-08-05. Population: 51,201 tasks across 1,375 users, demo account excluded. Auto-generated recurring instances, 29,546 tasks or 57.7% of the raw count, are excluded from every figure, leaving 21,655 one-off tasks. Sub-samples are stated inline where they differ: 7,506 completed tasks that had a planned day, across 853 users, and 340 users with at least 5 one-off tasks.

Four limits, in descending order of how much they should bother you.

A task app cannot see work done elsewhere. This is the big one. Plenty of these incomplete tasks were done in real life and never ticked off, because the person did the thing and got on with their day. I am measuring the list, not the person, and every completion rate here is a lower bound.

This is correlation, not an experiment, and the population is self-selected. Nobody was assigned to a condition. People who install a tracker are not a random sample: they already write things down, which cuts both ways, more organised than average and more prone to over-capturing.

Survivorship pulls the aggregate down. Users who abandoned the app entirely still have their unfinished tasks counted. That is honest, since dropping the quitters is how you manufacture a flattering number, but it means the 35.9% headline includes many one-visit lists nobody intended to work through. The 54.5% median among engaged users is the fairer comparison.

Rounding and scope. Everything is reported to one decimal place, no individual user or task text is included, and nothing is extrapolated. This is one app's data, not a claim about humanity.

The one sentence version

If a task is not going to move in the next week or so, writing it down more carefully will not save it. The only lever with real evidence behind it is writing down less.

I have never met anyone whose problem was a to-do list that was too short. Including me.

Written by Eusebiu, a developer of nine years, now building Loggd full time. The overdue queue in this data includes my own.

8 Months of Revenue, the Boring Version: €14 to €1,133 to €461

Eusebiu Balan — Thu, 06 Aug 2026 20:24:08 +0000

Originally published on my blog.

€14, €161, €145, €85, €522, €1,080, €1,133, €461.

That's 8 months of revenue. That's my whole run so far.

I see the other posts every day. 10k MRR in 30 days. Six-figure launch. 50k in a quarter. Those posts are why I started....but 8 months in, this is what building a product actually looks like for most of us, and almost nobody posts the boring version.

Quick context: I'm a dad, married, 8 years as a dev. I started this before work, after my daughter went to sleep, and every weekend. Two months ago I quit the 9-5 to try this properly. Part living the dream, part just needing out of the burnout.

The app is Loggd, an all-in-one personal growth app. Habits, tasks, goals, focus timer, all tied together with a GitHub-style activity graph for your life. Built solo on Laravel and Vue, one codebase across web, iOS and Android.

Where I'm at:

6,286 users
70 paying, 4 on trial
€3,621 total revenue
~€205 MRR
~150 DAU
~23 new users a day (August just started, so ignore that tiny last bar on the chart for now.)

Now look at that first row again

€14, €161, €145, €85. Four months. €405 total. One of them went down. I was still working full time and building at night for all of it.

Then April: iOS launch plus my first batch of lifetime deals. €522. Then €1,080. Then €1,133.

If I only posted that stretch it would look like a rocket. It wasn't.

The lifetime-deal trap nobody warns you about

Of the €3,614, about €1,800 came from lifetime deals. Half my revenue is money that will never repeat. In July I didn't run any and it dropped to €461. Nothing else changed. Same posting, same product, more users than ever.

Lifetime deals are great for cash. You feel it right away. But then you hit a quiet month and you really feel that too, because the thing underneath was never as big as the good months made it look.

That's why my MRR is €205 and not €1,000. A good month and a growing business are not the same thing, and I only learned it by watching my own chart.

The other expensive lesson

I wasted about €1,400 on Meta and Google ads at the very start, before I knew anything. ROI was bad, I stopped. Expensive, but at least it was early.

A few things I'd tell someone at month one

The first four months are supposed to look like nothing. Mine did.
One-time revenue feels amazing and teaches you nothing. Watch MRR or you'll fool yourself.
6,286 users, 70 paying. Barely over 1%, and that's not good by any standard. The catch: web converts worst and has been live longest, so it's piled up dead signups. Mobile is newer and converts better.
Growth isn't a line. It's silence, a spike, then silence again.
The skill isn't writing posts. It's posting on the days nothing happens.

I'm not saying don't do this. Quitting was the right call for me, even at €205 MRR, and I don't regret it. Working on something you actually care about until 2 am...feels better vs being burnt out on someone else's roadmap at 3 pm.

I'm saying be ready for that first row of numbers. That's the part nobody screenshots.