Filters - desktop - A/B test

Hypothesis: Offering the customer the option to filter the full availability of car hire will increase conversion because customers need to be able to refine their search based on their wants and needs.

What I did: User testing revealed that customers wanted a way to filter the list of cars returned (the MVP was launched with just sorting options). Data analytics from other partners also gave a good gauge on what filters are most used, bring this info into the design.

Firstly prototyping to get the concept internal validation early in the process, taking on any feedback into the iterations. Open, closed, clear all states were considered.

After getting group consensus on the new filters feature I designed the UI with interaction states, then A/B testing this on a small section of traffic to see how customers interacted.

Results: This A/B test worked very well. We found that 25% of customer used filters and had a 54% increase in conversion when they did. Filters have now been rolled out to the full traffic.

Further iterations: I would like to test other filters for the number of people and to add other filters like extras (eg. baby seats). Further user testing will reveal further iterations.