An analytics team runs a customer-lookup application against a 4 TB table. The most common query is a highly selective equality filter on a high-cardinality column (customer_email) that returns only one or two rows. The table is clustered on transaction_date, which is used by a separate reporting workload. Query Profile shows that the point-lookup queries scan almost all micro-partitions despite returning few rows. Which optimization should you apply to speed up the point-lookup queries without disrupting the existing date-based clustering?