In the previous chapters, we explored different ways to optimize and save costs on compute resources. This chapter focuses on storage optimization techniques. Snowflake has a compute and storage decoupled architecture, which makes it flexible to scale and charge both resources independently. While compute costs are often highlighted, storage can become a silent cost burster if not properly managed. In this chapter, we explore strategies to trace unused storage and design considerations that can be crucial in curtailing the underlying costs right at the starting line.