Ending the daily "what should we eat?"
My family spent too long every day deciding what to eat — while dieting together. Meal Gacha! gives each family one shared spin a day. Changing the result costs a reroll, and rerolls are earned by working out.
The goal isn't finding the best restaurant. It's ending the debate — in a way that supports the diet.
So the decision belongs to the family, not to each person, and re-deciding should cost something you earn by moving.
Two rerolls to start. Scarcity makes the first spin matter; earning ties food to how much the family moves.
| Question | I chose | Why |
|---|---|---|
| Identity | 4-digit PIN per member | Accounts kill family adoption; open profiles let anyone spend your rerolls |
| Verification | Simple AI vision check | Peer review adds social friction; a yes/no check is cheap |
| Gacha rarity | Cut it — equal odds | Adds game feel, no decision value |
| Groups | Invite codes; one phone, many families | Friends wanted their own food pools; people eat with more than one group |
| Data | Screenshots expire after 7 days | Hundreds of photos add cost and privacy risk; a fingerprint is enough |
Any economy invites gaming. Testing surfaced each hole, and each got a specific fix:
Friends started using it and asked for things I hadn't planned: their own food pools, joining more than one group, and redeeming with Duolingo screenshots instead of workouts. The first two shipped; the third is next — pointing the product toward any healthy habit, not just exercise.
Watching real use also moved the new-member flow to the centre of the screen after people missed a small join button.
An AI coding agent wrote the code. I owned what to build and why: requirements, every screen review, end-to-end testing and launch on my own domain.