Skip to the articles
WRITING

On actually following through

Sorted by the way it usually goes wrong, not by date. Find the one that describes your week — the reset after one bad day, the tick you awarded yourself, the miss nobody saw — and start there.

14 pieces Last updated · August 14, 2026 Corrections · [email protected]
Standards

How we write these

Three rules, because this category is full of invented statistics.

  • 01 Every number has a source you can open If a claim cannot be verified it gets dropped, not hedged into vagueness. The accountability statistic everyone quotes cites a study that does not exist, and we say so.
  • 02 Competitors get their strengths stated Every comparison names what the other tool does better and links its own documentation, so you can check whether we described it fairly.
  • 03 We say what Commit isn't iPhone only, public beta through TestFlight, not on the App Store, and the published charitable-giving figures currently read $0. Pretending otherwise would make everything else here worthless.

Word counts on the cards above are body copy — everything in the article minus the breadcrumb, the updated line, the disclosure strip, the contents chips and the link list at the end — rounded down to the nearest 50. The exact figure is in each page's structured data.

Found something wrong? [email protected] — corrections get made, and the page says when it was last changed.

The one number worth citing 11% of smokers offered a commitment contract took it up, and those offered it were 3 percentage points more likely to pass a six-month nicotine test — an effect that held in surprise tests at twelve months. Giné, Karlan & Zinman (2010), American Economic Journal: Applied Economics 2(4):213–35. Three points is modest and real; the durability is the interesting part.