Compound improvement calculator (1% better every day)

The compound growth formula behind it

Improving by a fixed percentage each day multiplies your level by the same factor every day, so after n days the result is (1 + p) raised to the power n. Getting 1% worse each day multiplies by (1 − p) instead. With p = 1% and 365 days: 1.01 to the power 365 = 37.78, while 0.99 to the power 365 = 0.026. The gap between the two paths is 37.78 ÷ 0.0255, about 1,481 times.

The days to double come from the same rule: n = ln 2 ÷ ln(1 + p). At 1% a day that is 69.7, so 70 days; declining at 1% a day halves you in 69 days.

Small changes in the rate matter enormously over a year: 0.5% a day gives 6.2× after 365 days, while 2% a day gives about 1,377×.

How to read the result sensibly

The idea was popularised by James Clear in Atomic Habits as a way to show that tiny, repeated choices add up. Treat it as an illustration of direction, not a promise. Skills do not grow by a constant percentage forever: beginners improve fast, then progress slows as you approach a plateau. Nobody becomes 37 times better at running or writing in a year.

What the model does capture well is the asymmetry. A habit that slips slightly every day erodes you just as steadily as a good one builds, and the two paths separate faster the longer you look. That is why skipping a single workout is harmless but skipping one every few days for months is not.

Use the table to set expectations. After 30 days of a new habit the improving line is only 1.35×, which can feel like nothing. Most of the visible gain arrives in the second half of the year, so judging a habit after two weeks is the most common mistake. A more practical version is to pick one measurable thing, improve it a little each week, and check whether the trend line, not any single day, is going up.