Why this works.+
Wanting it is not the bottleneck. The version that survives a tired day is.
53%
How often people who meant to act actually did, across reviews of health behaviours.
Start with the version that survives.
A habit set at thirty minutes needs a good day. A habit set at one page can survive a bad one.
Save something good for it.
One podcast, one album, the good coffee: it only comes out while you do the small thing.
A cue can help. It comes last.
If your day has one reliable moment, start right after it. If the moment moves around, leave it out.
Where these numbers come from
Sheeran (2002)
Across health-behaviour studies, people who intended to act followed through 53% of the time.
Read the paperGollwitzer & Sheeran (2006)
If-then plans had a medium-to-large effect on goal attainment overall (d = 0.65), pooling lab and field studies together.
Read the paperSheeran, Listrom & Gollwitzer (2025)
0.36 across all 642 tests, with publication bias their own Egger's test calls substantial: 0.35 after trim-and-fill, 0.15 after a Bayesian correction that still finds extreme evidence the effect is real. Behavior change 0.27; physical activity 0.14; field 0.27 against lab 0.49.
Read the paperBélanger-Gravel, Godin & Amireault (2013)
0.31 right after the intervention, 0.24 at follow-up.
Read the paper
If you want to go further: Atomic Habits, James Clear.