Agreed!
When we launched Torta Stand, creating the assets and coding the game turned out to be the easy part.
The harder work was figuring out what makes it fun enough to keep playing, how to give players a reason to invite friends, and how to make progression challenging while keeping the economy from collapsing on Day 1.
This is one of five dashboards I use to track different stages of the player journey. Those trends help us decide where to make changes, and we’re still tuning.
Every decision about rewards, difficulty, or progression gives us something else to understand about how people actually play.
As someone who runs a mobile game studio and is also heavily in AI, it makes me laugh when someone thinks they can just solo vibe-code a $41m/mo game.
It is 20% about the core gameplay and the rest is live ops and experience tuning for retention.
These game studios are more like quants or psychologists than devs.
FTUE, D1/D7/D30, Churn, Cohort analysis, winback, progression, cosmetics vs power, ARPPU, CAC/CPI, ROAS.
Everything is carefully measured and quantified over months and everything gets AB tested. Not only ads but small features in game.
To make a game as financially successful as this is more about heavy data analysis and an insane analytics + live ops backend.
You can vibecode in a single prompt a gameplay MVP, not the hundreds of overlapping systems and have them balanced and calibrated for profitability.
This would take years solo.