How Many Devices are Enough?
In today's fragmented mobile device landscape, developers want to support as many devices as possible. How many devices is enough to test on? How do you know this?
The world has changed since I last delivered a talk on this subject in 2018 (https://www.youtube.com/watch?v=upqeuXS9N8M). Internet traffic from mobile device has doubled between 2015-2022, reaching 60% of all Internet traffic. The app economy is real as consumers have an average of 80 apps on their phones, resulting in a projected revenue of $935 billion in 2023.
These metrics come courtesy of mindsea (https://mindsea.com/app-stats/). How many apps do you have? How many do you make purchases in? Ok, there’s a lot of apps generating a lot of money…as an app developer, how do I test my app to ensure the best possible experience to get as large a piece of that pie as possible? The number one reason why people uninstall apps is they don’t use it. If it doesn’t work properly or crashes, people aren’t going to use your app. If they aren’t using it, you aren’t making any money. Now, on to devices themselves. While Apple has a relatively small number of iPhones in the market, there are a seemingly infinite number of Android devices worldwide. Outside of North America, Android is prevalent.
So why not just try to test on every permutation you can find? You probably don’t want to do that. Cloud vendors will tout their massive device farms and easy access, and this is definitely a huge benefit. The inclination is more, more, more. More devices, more tests. More complex tests. This may or may not be the best way.
During this session, Dan will cover the following topics:
• Using analytics to drive testing decisions. There are many statistics available today what platform someone is using your app on, how can you best leverage that information?
• Writing atomic test cases to get the most value out of test automation.
• How testing frameworks can shape your parallel testing strategy. Should you use Appium, XCUITest, Espresso, or something else? When should you use each of these?
Follow or Contact Us