As for your question in implementing indexes, you must know that the first step towards monitoring, and intelligently responding to index usage needs, is to collect and maintain a simple and easy-to-use data set of index usage metrics.
This data set should allow you to quickly search for common scenarios in which you might consider removing or altering an index:
- Unused indexes
- Minimally used indexes
- Indexes that are written to significantly more than they are read
- Indexes that are scanned often, but rarely the target of seeks
- Indexes that are very similar and can be combined
SQL Server provides a dynamic management view that tracks all index usage: sys.dm_db_index_usage_stats
.
This view is a cumulative total of operations against indexes and is reset when SQL Server services are restarted.
To learn more, please refer to this article https://www.sqlshack.com/sql-server-index-performance-tuning-using-built-in-index-utilization-metrics/