As a data engineer at a large ecommerce company, your customer order data is stored in Bigtable. The current garbage collection policy deletes data after 30 days, and the number of versions is set to 1. However, data analysts occasionally encounter customer data that is older than 30 days when running queries to report total customer spending. Your task is to ensure that analysts see only customer data that is no older than 30 days, while also minimizing both cost and overhead. How would you achieve this?