As an integral part of your data lake architecture, you plan to utilize Google Cloud Storage to store objects from various external systems. Each object will be ingested exactly once, and subsequent access to these objects will occur in a random pattern. Your goal is to reduce the costs associated with both storage and retrieval of these objects while ensuring that any cost-saving measures remain imperceptible to users and applications interacting with the data. What steps should you take to achieve this?