Improving search relevance is difficult. Learning to Rank (LTR) is an important and powerful technique using supervised machine learning to address the problem of search relevancy. A LTR approach leverages machine learning to automatically tune relevancy factors, which not only alleviates the pain associated with manual processes like boosts and blocks, but also promises significantly improved relevancy with the use of state of the art modeling techniques. This guide will demonstrate the power of the Fusion platform by combining LTR with insights derived from signals.