# Case Studies - Which query should be optimize first?
![`smart`](TiDBMonitorSQLPerf0401_en_001_20240125.png)
![`smart`](TiDBMonitorSQLPerf0402_en_001_20240125.png)
![`smart`](TiDBMonitorSQLPerf0403_en_002_20240125.png)

+ Query <B>
  + Only executed twice
  + Both the `Mean Visible Versions Per Query` and `Mean Meet Versions Per Query` are `17.7K`


------------------------------------------------------------------------------
In example `<B>`, the `Mean Latency` is 86.4ms, and both the `Mean Visible Versions Per Query` and `Mean Meet Versions Per Query` are `17.7K`, the highest among the three statements. However, with a `# Exec` of only two, this SQL statement is also executed occasionally, suggesting that optimizing it may not necessarily resolve system performance issues. Consequently, this statement should not have a high priority for optimization.

