Snowflake Search Optimization Service is specifically designed to improve performance for selective point lookup queries, which involve retrieving a small number of rows based on specific filter conditions. This is confirmed by the official Snowflake documentation referenced in the community discussion and the unanimous consensus (100%) among community members who selected option C. The other options are less suitable: A (queries against large tables with frequent DML) may not benefit as much due to maintenance overhead, B (tables larger than 1 TB) is not a specific query type, and D (queries against a subset of columns) is better optimized through columnar storage or projection pruning rather than search optimization.