Fernando Seminario - Senior Designer | Streamlining result cards on Hotwire
17699
portfolio_page-template-default,single,single-portfolio_page,postid-17699,ajax_fade,page_not_loaded,,vertical_menu_enabled,qode-title-hidden,qode-child-theme-ver-1.0.0,qode-theme-ver-11.1,qode-theme-bridge,wpb-js-composer js-comp-ver-5.2.1,vc_responsive
Product Design

Streamlining result cards on Hotwire

Hotwire is an online travel site offering Hotels, Car rentals, Flights and Vacation packages. Their bread and butter comes from a product type known in the industry as Opaque. The customer doesn’t know the name of the hotel or the brand of car rental until they book – basically a mystery deal. Here is my story about streamlining the result cards for the car rentals product.

Observation

Users are presented with a high number of car rental results – often over 100. Purchase data shows that most users don’t select a car rental beyond the 8th result on the page. Also, user testing reveals that users often struggle trying to compare different results when presented with so many options.

Hypothesis

If we only show essential information on the car result card, we will make it easier for the user to scan through the results and make a selection and therefore increasing conversion.

Insights

There was a lot of unfold so understanding the problem was important. I began by asking my product managers if we knew what content created value for the user. Some content had been tested in but not all. In order to facilitate this process with the team, I indexed all the content and gathered data from previous tests – some from all the way back to 2016. This helped later in the project by allowing us to pay special attention to highly valuable content. The list below shows test results for content that was either a winner, loser, or inconclusive. 

Research

A large study had recently been conducted by the car research team that included 3000+ participants. One of the insights was a list of the top 10 factors in the user’s decision making process when it came to renting a car. By leveraging this data, we were more able to closely align the design with existing user behavior in a meaningful way.

 Top 10 factors:

  1. Price
  2. Location of rental agency
  3. Rental car size
  4. Time it takes to pick up rental car
  5. Rental agency mileage policy
  6. Location of rental agency in airport
  7. Cleanliness of rental car.
  8. Rental agency cancellation policy
  9. Ease of understanding pricing on website.
  10. Rental agency hours of operation
No segment considered the exact name of the rental agency to be a top 5 factor.

– Research insight

Information architecture

In order to design a layout that closely aligned the user’s behavior, I looked at the problem from an information architecture standpoint. Three things were considering for this part of the process:

Heat map
User’s anchored to the price block, then car type (compact or minivan), model information, and car image.

Top factors
There was some contextual problems. The location content was in two different places (address vs. pickup type). This content ranked second in the research list of top factors so grouping this would save the user some time by eliminating the need check two different places of the design.

User testing
The car image only became a factor after users had validated the price and car type.

Height analysis

The design team had been making comments about the height of the cards for several months – saying it was too tall. I agreed to some extend, since for the past few years we had just been productizing winners that had small changes but didn’t look at the layout holistically. I believe that content dictates many things including height. The analysis revealed that our top competitor had one of the tallest cards – about twice as tall as ours! Go figure. Turns out we were average. No more comments =). 

Explorations

After collaborating with my product managers, it was time to move the design further along. Beginning with a solution to validate the existing content and then ending with a re-design. It’s worth mentioning that as a team, we agreed to attempt breaking down big changes into small testable chunks – whenever possible.

Exploration 1

This exploration focuses on validating the existing content on the result card with some minor styling updates. The idea here is that by removing content that isn’t creating value for the user, we can remove distractions and make it easier for the user to find what they are looking for. At the same time, setup the groundwork for the transition into a re-design.

Proposed test plan:

  1. Customer review content
    1. Hide thumbs up removal
    2. Hide “Recommended”
    3. Hide “/10” from rating
  2. Location content
    1. Reposition pickup type
    2. Airport name i.e. SFO
    3. Add icon to retail location address
    4. Change model text and distance text to grey.
  3. Value proposition content
    1. Hide “No credit-card required”
    2. Hide “Pay now and save”
    3. Hide “Pay at pickup”
  4. Use “Hotrate” logo vs “Hotrate Car”
  5. Hide carry-on
  6. Remove bold from “Free cancellation”

Control

e1.0 Content validation only

Exploration 2 – Thats a good dog

This exploration primarily focuses on using only one selling proposition vs. several e.g. “Pay now and save” or “No credit-card required.” Then taking looked at using a “Dog ear” treatment for savings copy that the hotel team had just started productized as part of their redesign. The card style helped each result stand out from the other; albeit, at the cost of having less space (narrower width).

e2.0

e2.1 Card style

e2.2. Big car image

Exploration 3 – Switch-a-roo

This exploration respositions the car type information with the car image, aligning with the top factors from the research study and user testing. Additionally, a horizontal layout for the value propositions was explored resulting in more efficiently use of space (going horizontal instead of vertical) and improved scalability.

e3.0

e3.1

e3.2

Exploration 4 – Spicy

This exploration looks at the problem from a new perspective. Taking a streamline approach not for results themselves but also at list level. Influenced by analytics data showing that the sorting and car type filters are the most popular ways of narrowing down a search. Additionally, the top factors from the research study were considered by anchoring each section by the top 3 factors.

Design

After sharing the explorations in multiple design and stakeholders reviews and getting everyones feedback, it was time to move on the design phase. Phase one consists primarily of content validation while phase two is what the future vision. 

Design (Phase 1) – Content validation
This approach sets the baseline of a value driven foundation. It also helps future state of design by allowing the team to make informed decisions about what content to change. Lastly, some elements from the redesign were introduced.
Design (Phase 2) – Vision

Remaining re-design elements are introduced in small test groups. New features were added such as the customer review score, green 1-10 element that also aligned with our design system. Additionally, a new item was added to the value prop row highlighting the feedback from our post-booking survey i.e. “Friendly staff”. Lastly, zoning and content hierarchy styling was established. User testing is in progress along with eye tracking tests.

e2.1 Card style

e2.1 Card style

e2.1 Card style

Content zones Once the design was done, it was important to establish some guidelines and zones for content in order to avoid ending up back where we started. Each content type has its zone and will be considered for future designs.

Primary content guidelines Aligning with the research study and user testing, this style guideline establishes that primary content should be easily recognizable to the user. This is mostly done with a slightly larger font size and thicker weight.

Secondary content guidelines The secondary or tertiary content will use a more subtle color, allowing the primary content to get more recognition.