Tedious, repetitive, and manual tasks are the first candidates for automation in a process. However, the automation itself requires time, money, resources, and personnel. So, it is key to understand whether the Return on Investment (ROI) of implementing the automation will be more beneficial than continuing to perform the tasks manually. Think of the most basic test involved in an application. Assume that it requires two teammates and six hours of time from each of them. If, by automating the test, we can reduce its execution time to only one hour, with only one collaborator in charge, then automation should be considered a priority (one added value would be the ability to reuse the test automation for testing similar applications in the future).

Continuing with the preceding example, what if implementing automation for the test will take a month, and will require a team of five testers? What if it is unlikely that we will be able to reuse the automation project in future applications? What if the test is so complex that the resources required to maintain the automation script will be considerably higher and more expensive? Depending on the answers to these questions, automation may not be the best option. There will be cases where analyzing the situation will not be as straightforward as in the cases mentioned previously.

The general process of deciding whether to automate a system should involve considering the answers to the following questions:

• How many resources and how much time are required to build the automation solution?
• How many resources and how much time are required to maintain the automation solution?
• How likely is it that the solution will be used in the middle and long term?
• What is the risk exposure level if the system isn't automated? (Even if the implementation of automation is extremely expensive in terms of resources and time, it might help to mitigate huge risks for the web application, in which case, it would make sense to implement it.)