In the high-velocity world of local commerce, the bridge between a user’s intent and a successful transaction is built on the precision of real-time advertising systems. This talk explores the "Personalization Flywheel" : an integrated ecosystem of retrieval and ranking architectures designed to optimize the entire user journey from search to conversion. We deep-dive into the technical evolution of the Ads Quality systems in the industry, focusing on two critical breakthroughs. First, we examine the transition from traditional keyword matching to Semantic Search, sequence models, and Embedding-Based Retrieval (EBR). By implementing advanced query parsing and enhancement techniques, we successfully unlocked the value of high-intent branded searches, resulting in a xx% revenue lift on some surfaces and significant improvements in Return on Ad Spend (ROAS).