What is the consequence of exceeding the recommended number of columns per clustering key?

• A. It improves pruning
• B. It increases costs more than benefits
• C. It has no effect
• D. It reduces latency

Answer: (B)

Clustering keys work best when you keep the number of columns relatively small. Each extra column adds more data combinations that Snowflake has to organize, track, and maintain for clustering. While a clustering key can improve pruning by narrowing which micro-partitions are scanned, the benefit grows only up to a point. When you exceed the recommended number of columns, the incremental pruning advantage tends to plateau, but the overhead of maintaining and reclustering increases. The system may spend more compute and storage effort on metadata and maintenance tasks than on gaining meaningful pruning, so the overall cost grows while the performance benefit does not scale accordingly. In practice, this means exceeding the recommended column count tends to increase costs more than it improves query performance.