First approach to Mobile app testing on NNOT – Noordertest Groningen (NL)

Today I gave a new presentation about how-to coördinate all task to test a mobile app on the NNOT conference. NNOT is small conference for IT departments from Dutch government organisations in and a round Groningen.

There is a lot of information and training material about what a tester can or should do to test a mobile app. This is more on an personal level of a tester. But how to organize the testing, and which tasks coördinated together? The role or the tester is different, more facilitating than doing the test himself and there are different task to do like creating personas and analyse the device market. But the most crucial task is analysing the conditions the mobile app should run in and translating these conditions into perspectives like network conditions, ergonomics or usability.

The full presentation of the NNOT is here.

City Groningen, Netherlands

NNOT presentation coordination of mobile app tasks

Mobile app Europe conference 2014 – Learn how to create better mobile apps!

Mobile app Europe conference 2014 – Learn how to create better mobile apps!

If I would sum up the first Mobile app Europe conference at Potsdam in Germany, then it would be don’t look at mobile first, but only make it for mobile. As Julien Lesaicherre from Facebook said: “Mobile is eating it all up, web is dead”. This is a reference to the Wired article from 2010 (WIRED – The Web Is Dead. Long Live the Internet and WIRED – The Web Is Not Dead)

So if it is not only mobile first, but in the end also be mobile only then what should we learn to create better mobile apps?

By putting the mobile first the focus is on the core functionality and only way a user is going to interact with a business. The mobile app need to start small but really stands out of the crowd. Then with each iteration learn from the feedback and build up the total platform. If mobile is the second channel the change on success is smaller because the organization or development team want to resize all the functionality to fit on the smaller screen. If you look at the mobile app, then you can see it has a complete different user experience and uses complete different technology. A PC is used in a work environment with perfect connectivity and ergonomic settings, has a big screen with a keyboard and mouse. Mobile devices are used constant around you, even in the toilet and bedroom. With a minimum of functionality you want the perfect mobile app user experience in all conditions for all users and all emotional states the user is in.

The conference can be summed up in the following points: think about the bias that everybody has, use the emotions that a user also has when he or she uses your app, test as the users by becoming them (impersonate) and in their environment, meet every week a number of users and develop the users (personas), find out what they need and not what they want or ask, wearable’s are going to be big and more personal because you where them on your body, fragmentation is a given for every tester and with a good strategy you can cope with this, performance, security and privacy are always in scope for every app, look at the default test set from Aqua they can help you from making stupid mistakes. Last but not least the robots are coming and they can also be used during testing.

Judging People – Why We Are Wrong When We Think We Are Right”

This is the title of the presentation from Dr. Chaehan So. For me it was the best presentation, because it involved the core of what testers do in their daily life: create an experiment and judge on what they see as output. But what are you seeing is influenced by your brain. Our brain is not neutral. What happens with you if you think you’re are right, but you are actually wrong? How do you feel? Good or bad? What happens if you are told you are wrong, if you thing you are right? Do you feel good or bad?

Misconception of psychology:

  • Trait versus behavior. Psychology start with traits, characteristics, and then monitor its behavior. You can’t determine the characteristics based on the monitored behavior.
  • Your brain is searching for evidence of confirmation and not for disconfirmation. Bias to neglecting information that undermines the idea
  • Your brain has selective attention, it is searching for the positive feedback
  • Your brain has a selective interpretation, it will filter out the negative ones and only show the positive feedback.

This means that if you see different behavior you can’t say there are different characteristics. You need research, experiment with settings and then monitor it change in behavior. If you think your are on the right path, you get in a certain thinking path and it is hard to get out. Also because your brain is pushing you that direction. A way out is by rational looking for alternatives. Accept that you make mistakes. Your ideas are based on assumptions not the truth . If you allow yourself to make errors to validate your assumptions, you stay open for different possibilities. The presentation aligns with the messages from  Gerald M. Weinberg:

If I can’t think of at least three different interpretations of what I received, I haven’t thought enough about what it might mean.

If you think you are right you feel good, real good. Sometime as a tester I feel like making a little dance. It can take hours to find a bug and then if you find it you feel good, real good. If someone comes to you and tells you, you are wrong. It hearts your self esteem and can lead to a self threat spiral. You feel bad so ignore the message bounce back to your feel good mode. You need to stay open minded. To do this you need your team around you. They can give feedback on the ideas that you have. The better the process is of self evaluation, best on good feedback, the better knowledge you gain from your environment. I this way you can better decisions. Dr. Chaehan So has put forward his PIMP model. With this model you can give feedback to overcome this self esteem and self threat spiral. Pimp stands for Perception-Impact-Proposal model. Give feedback that stands in relation to the goal you want to achieve. <add example of this model>.

<Example of the pimp feedback>

Fragmentation of the mobile world

While the presentation was not so perfect the theme Saurabh Agarwald put forward is very important. IF we want to test effectively a tester need to cope with the fragmented mobile world. How can you tackle this fragmented mobile world? He gave some strategies, but the picture was for me not complete. If I add my experience then I see the following options:

  • Use real devices that are in your local test lab
  • Use real devices in the cloud
  • Use a local open device lab
  • Use a crowd sourced devices
  • Use emulator or simulators local
  • Use emulators or simulators in the cloud

With these possible options there is no strategy jet. Which option can be used when? Important input to choose an option are the a) risk analysis and b) the market analysis. So with the risk analysis the possible weak points from the products is determined and with the market analysis the target group is determined. The result of the fragmentation strategy is a matrix with all the parameters on one side, like brand and network conditions, and all the devices on the other side. Every iteration this matrix need to be updated to determine which devices in which option will be used. The point that Saurabh Agarwald not stressed out enough is that you need to have a local test lab with multiple devices and then be flexible. If you don’t have a device then you can buy it or search it in the cloud. The set of devices to test on can change every week.

For extra info see:

The conference was big fun. Hangout and discuss with other testers, developers and designers that work in the mobile app world. I can’t wait for next year.

Mobile app Europe conference 2014 – Learn how to create better mobile apps!