The Content Optimizer runs an iterative Bayesian optimization loop that seeds testable hypotheses against a locked Rank Factor taxonomy, applies localized snippet or page edits, and measures rank improvements against a competitive set using LLM re-rankers.