The enterprise Web ecosystem is becoming increasingly complex, comprised of multiple disparate partner sites, data sources and Web services - all of which make up the new mashup world. And while the method of delivering enterprise applications may be changing, the users expectations for quality, dependability, and uptime have not. In this presentation, I will define the performance paradigm facing enterprise mashups, while answering the following questions: How does an enterprise validate a positive user experience when they don't control all the pieces in the service delivery chain? If there are problems somewhere in the chain, how are performance issues isolated and diagnosed? What are the important data points that should be monitored and how can that data be shared and best utilized across a web ecosystem?