Data Analysis
Last updated
Was this helpful?
Last updated
Was this helpful?
To answer two of my research questions I carried out a data-analysis. The data that was analysed belonged to the three portals of Peugeot, Citroën and DS automobiles. The reason I used this data is that there were no tracking records of the Employee Benefit Program. I used this data as a representative to draw conclusions that will help me improve the design for the Employee Benefit Program.
The two questions I wanted to answer were:
How do users behave during the configure/signing process?
How do users use the portal per device? (Mobile, Tablet, Desktop)
PSA provided their data studio to carry out this data analysis. To gather the data I needed, I exported parts of data to an excel sheet and filtered out all useless data. In this way, I got a clear overview of the behaviour during the configure process per brand and per device.
To visualise the data I used two variables, the pages and the bounce rate. The data, unfortunately, did not provide normal users flows, therefore I used the bounce rate as an indicator for page exits.
nb. Bounce rate is the percentage of visitors that leaves a website/page without clicking to the next page. Visitors that ‘bounce’ are likely to lack interest in the presented content.
The main goal was to find corresponding events that occur on all the portals, or just at one or two. The data showed the difference between the portals at all the configure steps.
Firstly, the version option has a high bounce rate at both Peugeot and Citroën configurators. A fourth of the visitors left instantly when arriving at the version option.
Looking at the two version option designs I conclude that both lack affordance. The Peugeot version option provides some information per category and even tells the price differences. The only problem is that not every category has an information bubble. Above that, the given numbers in each title do not provide an explanation as well.
The Citroën version option lack information in every kind. Only a name does not provide any information to users.
Insights:
Provide a clear explanation if using specific terminology
Provide price differences and reason between models
Another high bounce rate that occurred at Peugeot and Citroën was the car option page. Compared to DS Automobiles which had a bounce rate of 30%, Peugeot and Citroën lost almost two-third of their users at this page. To be exact, Peugeot has a bounce rate of 62,03% and Citroën 61,66.
These amounts occurred on desktop, the bounce rates on mobile and tablet were even higher for Peugeot. Tablet reaches 68,5% and mobile 80,51%. On the other hand, Citroën does it better when it comes to portable devices. The car option page reduces in bounce rate to 50% on tablet and 58% on mobile. Let us have a look at the designs.
When comparing the three designs it becomes more clear why the bounce rate is significantly higher at Peugeot and Citroën than at DS Automobiles. The options at DS Automobiles are structured in clear sections whereas Peugeot and Citroën use simple vertical alignments.
Thinking from a users point of view all the information and explanation can be quite overwhelming. Because the information is more structured users can stay focused on the particular section they need to answer.
Insights:
Answering questions which are hard to visualise should be well structured
Provide questions in large sections so users can stay focused on the particular questions
The landing of DS Automobiles lacks structure and feels unsafe. The first impressions says it all, while this landing is not inviting.
