Which column statistics should you update to improve query performance when table statistics are out of date?

Which column statistics should you update to improve query performance when table statistics are out of date?

• A. columns of the VARCHAR and NVARCHAR data type
• B. columns used in GROUP BY clauses
• C. columns with high cardinality
• D. primary key columns

Answer: (B)

When the optimizer doesn’t have up-to-date data about how a column’s values are distributed, it can misestimate how many groups will result from a GROUP BY and how many rows will fall into each group. This affects plan choices for aggregation, sorting, and intermediate results. Updating statistics on the columns that appear in GROUP BY expressions gives the optimizer accurate cardinality and distribution information for those groupings, helping it pick more efficient plans (fewer spills, better memory usage, and improved join/aggregation strategies). So the most effective targeted refresh is on the columns used in GROUP BY clauses.

When the optimizer doesn’t have up-to-date data about how a column’s values are distributed, it can misestimate how many groups will result from a GROUP BY and how many rows will fall into each group. This affects plan choices for aggregation, sorting, and intermediate results. Updating statistics on the columns that appear in GROUP BY expressions gives the optimizer accurate cardinality and distribution information for those groupings, helping it pick more efficient plans (fewer spills, better memory usage, and improved join/aggregation strategies). So the most effective targeted refresh is on the columns used in GROUP BY clauses.