The Habit Zone: Frequency Times Utility
Not every product should chase habits, and Eyal is explicit about the boundary. Plot two axes: how often the behavior occurs, and how much perceived utility it delivers over alternatives. Habits form in the zone where frequency is high enough, roughly weekly or more, even if utility per use is modest.
Infrequent products can still thrive, but on a different physics. Amazon purchases or tax software win on utility so large that users return deliberately. They're businesses of value, not of habit, and dressing them in streaks and notifications mostly annoys people.
The honest first question for a builder: is there a plausible weekly-or-better use case? If not, stop designing hooks and start designing for painkiller-level utility, because no loop survives monthly frequency.
Key idea
Habits need frequency: roughly weekly use or better. Low-frequency products must win on outsized utility instead, and no amount of hook design compensates for a use case that's rare.