komet.
The players who stay for months cost more to acquire, so an industry that needs its money back in a week stopped building for them. The games they want used to take a hundred people and three years. That part just changed.
-
01The audience is the opportunity.
Most puzzle players want to relax, not to be beaten. They are the larger half of the market and they pay for longer. Almost nobody is bidding for them, because serving them means waiting months for payback instead of days.
-
02The price of one attempt collapsed.
A serious attempt at this genre used to mean a hundred people and three years, which is why only giants took one. One person directing agents can now put a playable game in front of real players in weeks, and a failure costs a weekend instead of a studio.
-
03Meta is not the enemy. Competing metas are.
A coffee shop sells sofas, music and cake, and it is still about the coffee. The industry stripped meta out of puzzle games because the meta had been built to fight the core. We build one game, where everything points at the board.
-
04A pod is a studio and a publisher in one person.
One product engineer and five agents. The person who builds the game is the person who decides whether it lives. That used to be two companies arguing. AI is not the product; it is what lets one person hold both jobs.
-
05We kill loud, we kill early.
Most concepts deserve to die. Killing one before it gets expensive is the strategy working, not failing. Inconclusive is a result. Limping quietly forward is not.
-
06Scale is earned.
Signal decides what scales, not belief. Every stage has a pass bar written down before the work starts. When retention is real we commit fully: live operations, content cadence, the depth this genre demands. Before that, nobody is hired.