Honest.
Before we start.
Legal, the number one.
Prediction markets are regulated gambling in most Asian jurisdictions. China mainland prohibits them de facto. Korea criminalizes online betting. Japan regulates strictly. Hong Kong and Singapore have specific regimes. "It's not a casino, it's a prediction market" is a losing legal argument elsewhere.
Requires: legal opinion per jurisdiction (HK + SG + EU minimum), active geoblock from day one (including CN mainland IPs), tight marketing language, probably offshore legal structure with disclaimers, likely KYC for larger positions.
Technical & product.
Resolver corruption via collusion with disputant. Front-running on-chain. Admin keys and upgradability. API dependencies for data sources (Binance, Twitter, Weibo). The chicken-and-egg of liquidity. Category fatigue if a slow news week kills a vertical.
All manageable. All documented. None of them undocumented.
Narrative burn-out.
If prediction markets as a category cool off after Polymarket V2, the launch window closes. The MVP must land before Nov 2026, not after. Not "nice to launch soon" — a dated deadline, dropped if missed.