What criteria do you use to choose functions for automation and which are better left manual?

When choosing functions for automation, the following criteria are usually considered:

• Repetitiveness: functions that need to be checked frequently are better to automate.
• Stability: if the functionality is stable and rarely changes, it is beneficial to automate it.
• Complexity and volume: large and complex tests that are difficult to perform manually are better to automate.
• Risk of errors during manual testing: if manual testing often leads to errors, automation is advisable.
• Execution time: if manual testing takes a lot of time, automation can speed up the process.

Functions better left for manual testing include:

• New or frequently changing functions where automation may quickly become outdated.
• Tests requiring subjective assessment (e.g., UI convenience, visual design).
• One-time checks or very rare scenarios.

Thus, routine, stable, and business-critical functions are automated, while exploratory, new, and visual aspects are tested manually.