How to develop automation test scripts faster?

Automation testing is a proven way to quickly develop new features while ensuring the overall quality of the application. In small teams, it may be necessary to develop tests faster as features are developed.

Here's my strategy for achieving efficient Test Automation in the shortest possible time:

1. Be crystal clear about the test case or the flow that needs to be automated

   • A significant amount of time gets spent in these first few steps starting with understanding the test case clearly, clarifying every doubt discussing with the teams and the stakeholders.

2. Execute the test case manually a good number of times.

   • Through executing, the understanding gets clearer, observing behaviors of the application based on the actions performed, getting the feel of how a normal user will use it as eventually I am simulating a user.

3. Plan out the automation steps by writing pseudo code in my test file.

   • Based on the previous two steps, once I get an idea of the test case and how I am going to approach it, I write test steps as pseudo-code in the test file which helps me to free my mind and focus on each individual step at a time.

4. Create the Page Object Model and Test Data.

   • Once what to do and how to do is clear, it's time to get into the real action which is finding the locators. Mostly it's easy to do for me, besides when some weird elements come in, that's why I always have the documentation open. But overall I do try to finish this task as quickly as possible, and GitHub Copilot helps a lot. For test data, mostly it's provided by the manual tests and stakeholders.

5. Define necessary functions based on the actions and assertions to be done.

   • This is another place that sometimes requires a good effort so that the test can handle the dynamic inputs (test-data) and application behavior without duplicating code a lot. Here mostly I follow the coding best practices as it's the only way to go fast.

6. Implement test steps using test data and page objects.

   • Here the task is to just put all the puzzle pieces together, implementing the test steps using page objects and test data that has been created earlier. Sometimes I run the test cases when every step gets implemented, often times I implement group steps, make it work, and then the next steps.

7. Execute the test and Debug.

   • Finally looking at the final result and most of the cases it fails. The way to make it work is to debug and fix it. Having a solid understanding of the test case and the application behavior makes it really easy to point out the issue and fix it.