when do you really need to ditch the monolith for microservices

Weve been running a pretty standard monolith for years, and it served us well. But lately, with a growing team and more complex features, deployment cycles are getting longer, and individual service failures can bring everything down. Im starting to think about breaking it apart. What were the key indicators that made you decide it was time to move to microservices? Were there specific pain points that finally pushed you over the edge, or was it more of a proactive architectural decision? What did you get right, and what do you wish youd known before starting?