Recently, I was asked to present our approach in Mobile App testing and Mobile Testing in general for a panel discussion with different service providers. What I’ve noticed there is that almost all service providers directly went into the details on e.g. security testing. One of them came up with large scale solutions on security testing that will probably cost quite a few Euros. But, what if security is not really a big risk for your specific mobile project. You’d be spending a lot of money on security testing, while other and bigger risks are emerging for your specific project.
My main principle in testing has always been ‘No Risk, No Test’. Or in other words, always focus on the risks of the subject under test you are building. This should be a general rule in testing, not just for mobile apps and devices. The only difference that you can see in mobile is that you get additional risks when it comes to building new products.
- Interruptions like SMS, calls, network outage, battery removal, …
- The possibility of landscape and portrait mode
- The fact that mobile devices have motion sensors these days
- The status of the device – think about Flight mode for example
- Interactions between different apps – almost every app is currently having some kind of interaction with social media apps
- the installation procedure and possible interruptions in this installation like low battery when starting the download
- The battery status as such – most mobile devices limit the performance of the device automatically when the battery level is low
- Localisation of the app you’re developing – if you want customers in the US and Europe using it, this might require some adaptations
And besides from all this, there is of course the security risk. In our case of the grocery app, the fact that you can do payments in it means that your security risk is huge and will require a large testing effort before actually putting it in production.