Rentalcars.com – search results: reduce car type filter from 12 categories to 7.
The Brief
On search results the car categories filter is displayed as a carousel. From analytics it resulted that a good part of the users did not realise that they have to vertically scroll in order to see the rest of the car categories. This meant that the users did not have access to tailored information, could not find the right car and, as a consequence, could leave the website without making any bookings.
The agreed approach was to reduce the number of car categories from 12 to 7. The criteria for identifying the 7 new categories was: price, number of seats, number of luggage, special features. In addition to these, analytics played an important role in defining the new categories, selecting the ones with the highest results to be kept while incorporating the low performers into top performers.
By doing this I presumed that we eliminate an unnecessary step in the booking process. As such, users will see all categories at a glance and will not have to spend time struggling to find a specific category that can bring them closer to their desired car.
What I Did
- Problem definition. The first step of my process is to identify the underlying problem and to reduce it to something that addresses a critical human need. The core issue is that users are required to analyse a large set of content and exclude any results that don’t meet their criteria. If the users can’t interact properly with the tool and don’t have sufficient visibility and ease of browsing, then they will get frustrated and confused and will abandon their activity.
- Observation. Next, my process continues with observing and identifying the users who will use the filter. I get input from multiple stakeholders including the Product Owner. The next step is to synthetize the ideas for new categories and come up with a list of “finalists”. However I have to never lose sight that the user need is at the center of the problem.
At this step I look in analytics and verify if there have been any other similar experiments to identify possible outcomes that can impact the current task. Additionally, I performed a competitors analysis to identify how they are labelling and grouping their categories.
I get input from multiple stakeholders including the Product Owner and incorporate the relevant ideas to find the appropriate solutions. What I have to always keep in mind is that the user need is at the center of the problem and that is what needs to be addressed in an efficient matter. - Ideation. The last step is brainstorming. This is done based on what I found during the previous two steps: problem definition and observation.
I collaborated with the Product Owner and the Copyrighter to uncover possible insights. This leads to transforming concepts into features that address customer behaviours. This means that the label categories have been designed to provide a high level of familiarity to the user and address the human need of orderliness and succinct yet relevant information. During this step I also produced mockups and wireframes to illustrate the ideas.
- Prototyping. I developed a hi-fi prototype after receiving feedback on the draft versions. Additionally I also integrated HTML code.
Deliverables
- Mockups, wireframes;
- Hi-fi prototypes and HTML code. Please see bellow a deliverable example.
Results & Business Impact
Based on analytics, I concluded that users interact better now with the car category filter on search results. All categories are easily available and users don’t spend as much time as previously to find a specific car.
- After being A/B tested with users it became a gold experiment and is fully on now, increasing conversion with approx. 10%;
- Improved effectiveness as users interact more often with the filter;
- This version of filters performed better than previous with users in terms of CTR (Clicks Through Rate) as the car categories which the user had to scroll for had an increase in bookings.