Higher cardinality before lower cardinality reduces effectiveness

• A. It reduces effectiveness of clustering on the lower cardinality column
• B. It improves performance
• C. It has no impact
• D. It triggers automatic reclustering

Answer: (A)

Clustering key order matters because Snowflake prunes micro-partitions based on the combination of clustering keys. The leading column should be the one most useful for pruning in typical queries. A low-cardinality column is often more selective for many queries, so placing it first lets Snowflake narrow down to a small set of partitions early, and the remaining keys refine within that subset.

If a high-cardinality column is placed before the low-cardinality one, the initial pruning step is less effective. The first key splits data into many fine-grained values, which doesn’t align as well with common filters on the lower-cardinality column. As a result, you end up scanning more partitions overall, reducing clustering effectiveness when queries target the lower-cardinality column.

So, the statement that higher cardinality before lower cardinality reduces effectiveness reflects how the leading key governs early pruning. It’s not about triggering reclustering or guaranteeing a performance boost in all cases, and it does impact pruning efficiency.