In a world of increasingly faster software releases, our production environments are no longer static for a couple of weeks between deploys, they’re an ever-evolving, moving target. In this environment, accurate and informative monitoring is paramount. However can traditional monitoring practices fully provide this? We need to know exactly what our users are experiencing, as soon as possible, not merely machine-level metrics. As a solution, I’ll discuss the concept of Synthetic Monitoring; simulating user behaviour to provide monitoring ABSTRACT alerting. I’ll share my experiences of implementing this, and demonstrate how we can repurpose existing testing frameworks and tests to run against production systems as continual, user-centric monitoring services, that greatly enhance the accuracy and relevance of traditional monitoring