2. Assembling the team
Although testing automation heavily relies on software (and often hardware), it’s the humans that will set up the whole process, operate the hardware and software, and process the results. An automation QA team needs a QA Lead or Key AQA, who will make executive decisions concerning the project, and at least one Automation QA to perform daily tasks.
3. Gathering the requirements
Before the team can make any QA-related decisions, they need to know what they are working on. In other words, they need to get a complete picture of the project in front of them. This can include the specifics, what has already been done, and what is in the pipeline for the near future.
4. Identifying the risks
Like any new endeavor, introducing automation to a QA project comes with its own risks, so the job of the team at this stage is to identify those risks and propose solutions to mitigate them. For example, incorrect estimation of time and personnel needed to complete a project is one of the common risks, and one of the ways to mitigate it is to make the goals and estimations more realistic with the help of valuable prior experience and best practices.
5. Selecting cases for automation
This is an integral step of designing an automation strategy because it determines the scope of work for the automation team in the upcoming months. When selecting the test cases, it’s important to strike a balance between the desire to automate everything and the need to spend the available resources wisely. An experienced Automation QA will be able to easily choose the most appropriate cases for automation.
6. Choosing the framework and tools
The automation testing framework and tools will directly impact the efficiency of the team’s efforts and whether the project ends up meeting the goals set in the beginning. When choosing the tools and the framework, it’s necessary to take into account not just the needs and specifics of the project, but also the familiarity of the team members with the necessary technologies.
7. Setting up the environment
The test data and test environment are two aspects of an automation QA project that can make it efficient or slow it down in case wrong choices are made. This step typically includes setting up the equipment and preparing the software, establishing the schedules, and appointing the person who will be responsible for maintaining the test environment on a daily basis.
8. Writing the scripts
A group of well-written test scripts can give the project a sense of direction and help the team utilize the available resources and app functionality to the fullest extent. When writing test scripts, it’s best to break down complex scenarios into single-purpose tests and only then combine them into larger blocks if necessary. It’s also important to prioritize the scenarios, so that the most crucial elements of the application get the most attention.
9. Monitoring the results
After the start of the automation QA project, the team’s efforts should also be directed at monitoring the progress, analyzing the results, and finding room for improvement with the use of correct tools and metrics. This is done to prevent wasting valuable resources on tests that have little to no impact on improving the quality of the application being tested.
10. Maintaining the test suite
This is the part of the process that is sometimes overlooked by automation QA teams, usually to a detrimental effect. Neglecting to maintain the existing test suite results in the tests becoming outdated, far less effective, and wasteful of the company’s resources.
Popular frameworks for designing an automation testing strategy
The choice of the framework for an automation testing strategy is one of the most crucial ones to make, on par with the choice of tools and the composition of the team. Generally speaking, the choice of a framework depends on the specifics and end goals of the project. Other factors influencing the decision include:
- Basic functionality offered by the framework
- Availability of advanced features specific to the project, such as reporting, logging, debugging, and so on
- Types and levels of testing the framework can cover
- Familiarity of the team members with the framework
- Whether the framework is distributed for free or on a paid basis
- Whether the team has the skills and resources to modify the framework to fit the needs of the project
- How well the framework can be incorporated into the CI/CD pipeline
In many cases, QA teams prefer to create their own frameworks or adjust their previously developed frameworks to better fit the current project. However, there are also several popular frameworks that offer ready-made solutions for automating tests. Here are a few of such frameworks to consider.
Behavior driven framework
The behavior driven framework for testing automation will consist of a large number of feature files that contain scenarios based on the gherkin model (given/when/then) that are written in plain English. The key benefits of this framework are that the findings will be easy to understand even for non-technical users and that there are plenty of ways to reuse the code. The downside is that preparing the feature files takes a significant amount of time, and so does maintaining the cases.
Data driven framework
The key concept of the data driven framework is separating the data from the actual tests, so that they two could exist and be interacted with separately. This is a particularly suitable framework for projects where data is frequently changing — that way, the team can run the same tests with different sets of data. The downside of this approach is that collecting and classifying data takes time, while code reusability is not very high.
Modular driven framework
Under the modular driven framework, the entire test case will be divided into smaller modules, where each module includes a different set of features or other elements of the solution. The test case is then created for each module separately, and several test cases can be combined into a larger test suite. The modular driven framework allows the team to work in small increments and create test cases that are easily maintained. At the same time, breaking down the application into modules and creating separate test cases is a time-consuming and resource-intensive process.
The hybrid automation testing framework can be a combination of any two or more models described above. Companies often go for hybrid frameworks when dealing with unique features and requirements of a testing automation project. The decision to use a combination of several frameworks can provide additional flexibility, increase code reusability, and make room for more efficient reporting and analysis. The downside of a hybrid framework is that it can only be developed and used by high-skilled QA professionals, while a poorly designed framework can hurt the project in the long run.