Test Estimations
Measuring the complexity: Complexity derived from manual test cases, complexity of steps will be more complex in automation that manual steps(estimation based on manual steps)
Test data creations
Turn around time: for clarifications while automating(lack of knowledge in each and every step), Dependency on manual tester/developer
Rework: you might think some scripts/functions developed by colleagues will be useful, but when you script scenarios they might not be as per your requirement.
Monitoring the automation work as per the plan
Technical Issues
Technical support: Delay in technical support from the tool support team/lack of proper forums to communicate if any issue arises
Third party/Custom controls:
Object/Application changes: some object properties changes on the fly,
Tool limitations: limitations of installing extra softwares in client machines when there is work around with other softwares, tool bugs/patches
Compatibility: compatibility of tool, cross browser testing, cross platforms (win xp, win 7)
Intermittent application errors: You may see these in unstable environments
Lack of choosing rightframework
Choosing right framework: Each framework have advantages and disadvantages. We should foresee the expectation/changes and adopt a suitable framework.
Framework changes: When necessary change framework, which should not impact already developed scripts.
Lack of identifying suitable resources(human resources): unavailability of resources
Lack of process in identifying automatable test scenarios.
Rather than automating workflow/end-end flow including micro level test scenarios are not good for automation like when a dialog is opened, verify dialog title, click ok, cancel, short cut key etcc....
Here the intent should be maximum coverage with scripting.
Choosing the right tool for automation.
