Not defining Key Performance Indicators properly

Scheduling Performance testing at end of development and test cycle

Incorrect Workload Model

Failure to create a realistic test environment

Ignoring system errors

Conclusion