A Kafka stream feeds package-tracking events into BigQuery. The table was created partitioned by ingest date and query latency has increased. Analysts query geospatial trends by package. What change will improve BigQuery query performance?

Clustering the table on the package-tracking ID column will improve query performance for geospatial trends by package. When a table is clustered, BigQuery automatically sorts and organizes the data based on the specified columns. This allows BigQuery to efficiently prune data during queries, reading only the relevant blocks of data. Since analysts query by package, clustering on package-tracking ID will significantly reduce the amount of data scanned. Adding clustering on the ingest-date column is redundant because the table is already partitioned by ingest date. Partitioning already provides efficient pruning based on that column. Archiving older data to Cloud Storage and exposing it via an external BigQuery table adds complexity and doesn't directly address the performance of queries on existing, active data. Recreating the table partitioned by package delivery date would require significant data migration and may not align with how data is ingested or how analysts primarily filter their queries.