As a member of the SRE team monitoring production-grade applications, you observe a degradation in one application's performance, affecting customer experience. What is Google's recommended initial step in managing such incidents?