Why should you care?

After bulk loads, stale stats can make the optimizer pick scans instead of seeks.

See it live — copy this example

Run in SQL Server Management Studio (SSMS) or Azure Data Studio.

USE DataVerse;
UPDATE STATISTICS dbo.Orders IX_Orders_CustomerId WITH FULLSCAN;
DBCC SHOW_STATISTICS (N'dbo.Orders', N'IX_Orders_CustomerId');
SELECT name, auto_created, user_created, last_updated
FROM sys.stats s
CROSS APPLY sys.dm_db_stats_properties(s.object_id, s.stats_id) p
WHERE s.object_id = OBJECT_ID(N'dbo.Orders');

What happened?

• UPDATE STATISTICS refreshes the histogram.
• SHOW_STATISTICS displays density and steps.
• dm_db_stats_properties shows last_updated times.

Practice next

1. Update stats on Orders with FULLSCAN in lab.
2. Run SHOW_STATISTICS and skim the histogram.
3. Check last_updated for all stats on Orders.
4. CREATE STATISTICS on (City, Status) manually.
5. Compare estimated vs actual rows in a plan.

Remember

Stats drive cardinality estimates. Refresh after big data changes. Auto stats help; heavy ETL may need manual updates.

Post-ETL stats job

Nightly DataVerse load finishes then updates stats on fact tables.

Outcome: Morning reports pick sensible plans.