Test automation saves time taken for app version upgrades by 70%

Company profile

One of the largest furniture manufacturers in the world headquartered in Arcadia, Wisconsin, with distribution centers in Washington, California, New Jersey, and Florida in the United States, and global operations across Canada, Mexico, Japan, Central America, Japan, and Singapore.

From the first HomeStore in Alaska in 1997, the retail arm grew to become 800+ stores worldwide, and the brand is the #1 home furniture retailer in North America.

The challenges

1. A mobile app used by the retailer to enable e-commerce, and enhance the customer experience undergoes periodic upgrades.
2. During every release, the test case has to execute within a stipulated period of time, on iOS and Android platforms and a range of devices.
3. Separate backend API testing has to be carried out, to remove ambiguity between frontend and backend issues.

Solution

1. The use of automation frameworks for multiple platforms with a single code base has made it possible to cover all scenarios in a limited amount of time.
2. The testing protocol includes execution on multiple devices, and on different platforms parallelly.
3. The API testing in the application is also covered for each scenario to verify whether the response is received as expected and ascertain between frontend and backend issues

Some more important features

1. Implemented web browser automation on different browsers for both platforms.

2. CI/CD pipeline setup for scheduled automation jobs.

3. Modularized test cases.

4. Implemented a universal framework for mobile, web, and API testing.

Tools and technology

Android Studio, Appium, XCode, Simulators, JDK, IntelliJ/Eclipse, Jenkins, Github, Azure DevOps, TestNG/JUnit.

Adequate user training was offered to new hires, on the business process and wiki pages are also developed for future references.

Result

Decreased overall manual effort and number of man-days to execute the same suite on different devices and platforms.

Automation helps to execute regression suites for all the modules simultaneously. This improves the quality of the output and reduces regression testing duration by almost 70%.