Report performance can be improved through use of:

• Filters—Narrow the amount of data being processed through use of selective filters. For example, filter with specific dates or date ranges, table lookups, or other numerical values to improve response times. Use required pinned filters to ensure certain filters are applied before querying for data. 
• Flat list report layouts—Return a set number of records with a flat list layout. This is the preferred layout type for large amounts of data.
• Smaller groups—Group records based on less unique values. For example, if you group records based on company or org unit, you would have a limited number of groups. If you group based on transaction IDs, each record would have its own unique group. 
• Limited calculated columns—Select only the calculated columns necessary for your report. Having a large number may decrease efficiency. 

Performance considerations

Complex reports with large volumes of data may encounter performance degradations or time out errors. In general, you can expect the following results with different data volumes:

• 500,000 records or less—Expect records to be returned. 
• 1,000,000 records—Likely to be performant and return records, though complex pivots or row groups may cause degradation. 
• 3,000,000 records—Filtering your data is advised. 
• 5,000,000+ records—Highly selective filtering is necessary for performance. 

Complex pivot and row groups can impact performance. Consider changing your inputs if you experience any latency.