As part of continuous process of improving Volvo’s online car configurator I took the task of fixing the old problems while at the same time creating something completely new. Let’s take a look where we landed with the fixes.

This work started with an initial design system already in place, so we did very little hand sketching and just used the existing components to rapidly churn out designs. We also started to add things and use a more proper design system at the same time while moving to Figma.

Parallel design tracks.

At the same time, I did a parallel track on the whole revamp of online car buying journey. You can skip to that track by pressing this button.

To start the process we held standard GV design sprint with all the stakeholders. The idea was to pinpoint the problems the current site is experiencing when looking at analytics.

We traced the biggest problem to the very first page of the car configurator - selecting the trim level.

We already had a bunch of testing done for us regarding the page, so we had a starting point at least.

We did an assumption based on the research available; people don’t actually understand what the differences between trim levels are. The cars sort of look the same and the features sound a bit non-sensical.

Therefore our initial designs included a screen with giant imagery and a piece of copy explaining what the trim level is supposed to be.

“Our starting point was this extract from Tableau. It clearly shows that we had a problem with trim selection”

In testing this prototype, our carousel was not well received and people just blazed through it without looking at trims at all.

We conducted the testing using Userzoom and remote partcipants for the first 3 prototypes.

For our second prototype we mixed things up a bit. Trying to digest all the information we got from the user test our first goal was to fix the issues we found out.

Our second sprint was all about adding functionality to compare the trims against each other with a simple user interface.

We removed the carousel and put the different cars under each other so it’s easier to see the trims. We also put the price tag right there, so people would understand that they are getting more bang for their buck.

Finally, in the second selection page we included a “Compare trims” link for people to quickly compare the different cars.

Again, we tested 20 people remotely. It was now crystal clear for us to remove all references to ‘trims’.

Half of the respondents didn’t understand the word trim at all. And people still found that first selection screen more of a distraction than helping them select the correct trim level.

We removed all references to ‘trims’, the initial selection page, minimised the gallery to a thumbnail and condensed the information to be glanceable and as minimal as possible.

Here is a user testing video of the prototypes.

These are the conclusions after we tested these prototypes. At this point we just did guerrilla testing as it far more easier and faster.

I tested our prototypes against these hypotheses and described my findings underneath.

At the same time while we (UI designer and me) were creating and testing improvements to the old system, I started track that looked the car buying process from a customer perspective.

By using the data and insights we gathered thoughout the improvement track, Google’s excellent car buying process research, and our own data on which cars people mostly buy, we created 5 hypotheses to kick us off.

I created this flowchart to illustrate how this this system would work based on the synthesis of all the research we did over the months. Our main insight was that people don’t want to start with a blank canvas, but rather modify to suit their needs, hence we would start the journey by already having a car built and the price tag attached.

This was our end-goal, a car that has all of the decisions already made.

This simplifies the cognitive load for people on their customisation process by having the choices pre-made and seeing the price tag change in real time with each decision when changing little decision.

We wanted to be the digital car salesman, selling cars by the lifestyle rather than customers deciding what suits them through long-winded process.

Again, the design system we started with the design team made it so quick to iterate and create these designs on the fly.

The current site looks a bit different now that the new design system is fully fleshed out.