Several IT organizations invest in Performance "Testing" Centers of Excellence (CoEs) to cater to the needs and demands of their customers (internal and/or external). Testing, however, is only part of the performance story, and soon enough the CoE faces the business mandate to handle the broader area of software performance engineering. To be successful in meeting their business objectives and leverage their investments, the key imperative for these organizations is to have a clear definition of their existing and future CoE service profile along with a sound technology foundation.

However, even the mature IT organizations adopt a trail-and-error approach to defining, managing and executing the operations of their Performance CoE. Such an approach throws up perennial issues with respect to the following:

1. The precise nature of the relationship between the CoE and its consumers is often ill-understood, and the scope of the work and the deliverables that the CoE commits to is consequently ill-defined. The result is much finger pointing and firefighting that could have been avoided.
2. The core element of a scientific technology foundation in the area of software performance lies in some important quantitative techniques. This poses an entry barrier to practitioners, and there are few practicing subject matter experts available for help.
3. Industry-standard approaches for performance testing or benchmarking are restricted to the scope of a specific platform or technology. However, the bulk of any performance engineering approach or methodology is platform-agnostic and this is, perhaps paradoxically, not covered by industry standards.

So, what are the essential ingredients for setting up and running a successful Performance CoE? In this tutorial, the anchors address both the technical as well as the management aspects of this question using representative real-life scenarios and case studies.