Image showing crew member Pulse who is looking at the analytics for Swisscom's Kime portal together with robots Linz and Lang.

Analytics and User Experience: Swisscom's event portal as a success story

18.08.16

Using analytics, we can find out where UX enhancements are necessary—and where not. Doing so, we optimised Swisscom’s KIME event portal exactly where needed.

The development of the KIME event portal

For two years, Nothing Interactive has been responsible for the frontend of Swisscom’s internal KIME event portal. The portal was originally built as a minimal viable product, and iteratively extended. Now, after two years, it was time to enhance the overall User Experience (UX) of the portal.

Straight form the beginning, we took a user centred design approach to the KIME portal, and started with user testings to obtain qualitative data. However, it was not easy to get complementary quantitative data. For compliance reasons, Swisscom has to own their analytics data, so it was not possible to for instance simply use Google Analytics.

UX optimisation through analytics

For the last major release, a comprehensive optimisation of the User Experience of the KIME portal was planned. In order to validate our assumptions through quantitative data, we installed the analytics tool Piwik. Thus, we could track the interactions of the users—and gain valuable insights.

Validating assumptions

The KIME event portal allows users to create their own events. To do so, they record the details of the event in a relatively complex form. We assumed that this form would be an obstacle due to the cognitive effort which it requires. However, the quantitative data showed: only about 20% of the users aborted the process. A qualitative user testing confirmed that the logic of the form was well understood, while a visual refinement would be advisable.

Identifying potential

Furthermore, the event portal accounts for complex booking flows with real time availability calls, bookable hospitality services and integrated payment processes. As a result, the booking flow has multiple steps. At the end, users are shown an overview which they have to confirm to complete the booking. The analytics data showed that many users did not complete the process at this stage. Rather than confirming the booking, they navigated away, thereby cancelling the booking. Clearly, we had found a UX weak spot. We could then prioritise the issue and optimise the experience.

What can we illustrate with these two examples? With analytics, we can dismiss wrong assumptions. At the same time, we can isolate UX weaknesses which might otherwise go unnoticed. With analytics data, we can hence show where the User Experience should be optimised, and where not.