In January, for the first time ever, mobile apps were used more than PCs to access the Internet in the United States. Of the overall Internet consumption in the US in January, mobile-powered usage accounted for 55 percent, and of this, 47 percent of the traffic was from mobile apps. According to experts, these numbers will only increase.
As varied disciplines gear up for this wave, what does this mean for testers, who have traditionally tested web applications? Mobile application testing is not new, but what is new is the increasing set of challenges in this space by the day.
For example, there are very specific tests that a fast food app will need to run, such as user experience and load handling capacity at lunch times. In certain cases such as location-based apps, real-world testing is becoming inevitable.
Looking at this landscape holistically, mobile app testing in general differs in many ways from a standard web application testing process. With the increasing set of devices, it is a tough call for the testing team to decide whether to source and have all devices internally or to use mobile emulators.
In most cases, having the physical device is very important for the tests being run, at which time deciding how to group the devices to procure one to test from each group is another challenge. This may be a huge overhead up front, but if the team is trained to think along these lines, it will very soon arrive at an optimized strategy and be able to make a quick call on any new device with a considerable market share or potential that enters the market.
The challenges are alike regardless of your size of operations—for example, the Google+ team experienced trial and error before coming up with its meaningful testing strategies. The team gives a succinct list of unit tests, backend tests, UI tests, and monkey tests to run as a high-level approach to mobile app testing.
Another thing to closely watch for is the kind of application that is under test: a native, web, or hybrid mobile app. From there, you can decide what tools to use. In my experience, open source tools such as Selenium, Robotium, Calabash, and Appium provide a lot more flexibility and room for customization than commercial tools do when testing for mobile apps.
The scope and need for mobile app testing is only going to increase. Arriving at important decisions about what kind of app needs to be tested, how the app’s domain will determine what kinds of tests are needed, how to group devices in your testing matrix to arrive at an optimal set, what your device procurement strategy needs to be, and what tools you will use will help you build the much-needed differentiation in this space.
