The first evening with a new habit tracker is the most optimistic you will ever be about your own week. Water, gym, reading, meditation, journalling, no phone in bed, Duolingo, stretching. Eight habits, all reasonable, all things you genuinely want.
By the second Thursday you are doing two of them and feeling worse about your discipline than you did before you started tracking anything.
The problem is not discipline. It is arithmetic.
What a new habit actually costs
The reason eight fails is that a new habit is not free until it is automatic, and automatic takes a while. The University College London study that produced the widely quoted numbers found a median of about 66 days before a new daily behaviour felt automatic, with enormous variation between people and between behaviours.
Until it gets there, every one of those habits needs something from you every day: remembering it exists, deciding to do it, doing it, and logging it. That is four small acts of deliberate attention, and they do not overlap. Eight habits is thirty-two of them before you have had breakfast.
Note what is not the constraint here. It is rarely time — the whole list might be forty minutes. The constraint is the number of separate deliberate decisions you can reliably make in a day while also living your life.
New habits and kept habits are different things
This is the distinction that makes the whole question answerable.
A habit you have held for eight months costs you almost nothing. You do not decide to brush your teeth. It is attached to a fixed point in your day and it runs without supervision, which is what automatic means. You can hold a large number of those at once.
A habit you started on Monday costs you a real decision every single day. You can hold very few of those at once.
So the honest answer is two numbers:
- New habits in progress: two or three. One if the habit is genuinely demanding, like a daily hour of anything.
- Established habits being maintained: as many as you have earned. There is no useful ceiling. People run ten without noticing.
Most habit-tracker disasters are someone treating eight new habits as if they were eight established ones.
Picking the two or three
Sort the list by one question, not by importance: which of these, if it stuck, would make the others easier?
Sleeping earlier makes the morning gym plausible. The gym does not make sleep easier. Putting the phone outside the bedroom makes reading happen by itself. Reading does not do much for the phone. There is usually one habit near the root of the list, and starting there gets you two habits for the price of one.
Then apply a second filter that people skip: which of these can you do on your worst plausible day? A habit sized for your best week is a habit that will break the first time you sleep badly — which is the whole argument for designing a routine around the bad night rather than the good one.
The short version
Two or three new habits at a time, unlimited established ones. Pick the habit that makes the others easier, size it for your worst plausible day, and keep the rest of the list somewhere visible instead of active.
What to do with the other twelve
Deleting them feels like giving up, so do not delete them. Park them.
Keep the full list somewhere you will see it — a note, the bottom of the tracker, anywhere — and treat it as a queue rather than a backlog. It removes the fear that you are abandoning something, which is the actual reason people keep eight habits active while doing two.
Then add the fourth only when one of the first three has stopped requiring a decision. The test is not a streak length, it is a feeling: you notice you have already done it. Some habits get there in three weeks and some take four months, which is exactly what the UCL data would predict.
How to tell it is time to add one
Two signals, and you want both:
- You have stopped needing the reminder. If the notification still tells you something you did not know, the habit is not finished forming.
- Bad weeks no longer break it. One good month proves nothing. One bad week survived proves quite a lot.
The second signal is easier to read on a grid than on a counter, because a counter resets and a grid does not. A heatmap of the last fifteen weeks will show you whether the recent half is denser than the earlier half, which is the only trend that matters here.
Streaky lets you set a weekly target per habit rather than demanding all seven days, which is usually the right shape for a habit you are still forming. The streak tracker page covers how weekly targets and streaks interact.
Track it with Streaky
Free habit tracker with streaks, a heatmap and 24 achievements. No account, works offline.