Which technique is commonly used to improve Snowflake query performance by physically organizing data?

• A. Creating one or more materialized views (clustered or unclustered)
• B. Clustering a table
• C. Search Optimization
• D. Indexing

Answer: (B)

Clustering a table is the technique used to physically organize data in Snowflake to boost performance. Snowflake stores data in micro-partitions, and when you define a clustering key, you guide how the data is laid out to keep related rows together. This layout enables partition pruning: for a query that filters on the clustering columns, Snowflake can skip entire micro-partitions that don’t match the filter, reducing the amount of data scanned and speeding up the result.

Materialized views speed up queries by storing precomputed results, not by changing the physical data layout. Indexing isn’t how Snowflake works, since Snowflake doesn’t use traditional indexes. Search optimization isn’t about structuring the data layout for pruning. So, the most direct way to physically organize data for performance in Snowflake is to cluster the table, especially on columns that are commonly filtered or joined on.