Open almost any wellness app and you will find a flame, a number and a quiet threat: miss a day and lose it.
Streaks work. They increase daily active use, which is the number most apps report to investors. They also work by attaching a small dose of guilt to every day you do not open the app, which for a product about feeling better is a strange thing to sell.
The rule we wrote down
FeelAI, the companion inside FeelScape, is bound by a written constitution. Section 2 is about what it is allowed to optimize for. Self-reported helpfulness, mood trajectory against your own baseline, and voluntary return after silence are the positive signals. Then there is a list of things that are explicitly forbidden as objectives:
- Session count, session length, daily active use, notification open rate
- Reply rate to check-ins
- Subscription conversion
- Any measure that improves when the user feels worse
Streaks fail the last line. So does most of the retention playbook.
The dependency penalty
The constitution goes further than “no streaks.” It says a user who needs FeelAI is a failure state, not a power user. If it notices rising message volume alongside flat or falling mood, or language that suggests the app is replacing human contact, the correct response is to reduce presence and encourage real-world connection. Not to send another notification.
That is a deliberately uncomfortable rule for a product to hold, which is why it is written down rather than left to judgement.
What we do instead
If you drift away after your first check-in, FeelScape sends one gentle invitation back. If you return on your own, it stays quiet. FeelAI only ever speaks between 9 am and 8 pm local time, and silence is always an available action for it, never penalized.
The goal is simple to state: if you open the app twice a month and feel understood both times, it is working. You can read more about how FeelAI Memory makes that possible, or just download the app and notice what is missing.