The creation of high-quality speech applications is an art, and, especially for ASR, entails the tuning of the speech recognition performance. This tuning task is not a trivial one and requires specialist knowledge. The goal is to present many use cases taken from a wide variety of different applications of speech recognition in today’s speech market. This tutorial starts off assessing what makes a correct and sound tuning methodology and also covers best practices, suggested tips, and a trick.