when to prioritize eventual consistency over strong consistency?

hey everyone, im thinkin a lot bout consistency models lately, esp in distributed systems. we all know strong consistency is great for, well, strong consistency, but it often comes with latency and availability trade-offs. eventual consistency, on the other hand, can offer better performance and availability, but requires careful handling of stale data. what are your go-to scenarios for choosing eventual consistency? are there specific types of data or application features where you find its a clear win and how do you typically manage the complexities of potentially reading stale data in production