Date

4 weeks

Role

UX/UI Designer

Type

Website

Software Used

Figma

Date

4 weeks

Role

UX/UI Designer

Type

Website

Software Used

Figma

Date

45 days

Role

Visual + UX Designer

Type

Website

Software Used

AdobeXD

Project Overview

Background

I interned with Homegenius by Radian over the summer of 2023. Among the many projects I got to work on was the Room Condition Score Project. Homegenius is a network of home buying websites and products that leverage AI to enhance the home buying experience. One way this is done is by generating AI-driven room condition scores that rate the condition of different rooms in a selected property. In this project, D. (my teammate) and I aimed to explore the optimal ways we can display our room condition scores across multiple Homegenius platforms.

The Brief

The Problem

Initially, the Homegenius platforms used the 5-star rating scale to display the room condition scores. Due to the use of 5 star scales typically being displayed in tandem with user reviews, the concern arose that… 

...the 5 star scale did not communicate to users that the room condition scores were generated by AI - a key value proposition of the product.

The Goal

Research and define the room condition score display that best communicates objectivity and AI-derivation, and resonates most with users

Research & Empathizing

Foundational Research

I was tasked with initiating our research process by scanning the home buying industry and researching best design practices for rating scales. We started asking initial questions like:

    • What are user perceptions and expectations of room condition scores in various contexts (hotels, rentals, real estate)?
    • What perceptions do people generally have about AI-driven rating systems?
    • Which rating scale do users prefer: 5-star scale, numbered scale, or something else?

Findings

Here are our main findings from our research:

(click to enlarge)

User Persona

(click to enlarge)

Competitive Audit

D. and I conducted a competitive audit on 2 direct competitors that also used AI to help generate room condition scores. 

Our findings were:

  1.  Similarly to Homegenius, the AI-generated room condition scores correspond with the Fannie Mae property condition rating rubric
  2. Our competitors abandoned the 5-star rating scale and use a numeric scale and color-coding instead.

(click to enlarge)

Design Guidelines

After organizing our research, we decided on 3 main guidelines to inform our new design concepts…

Develop

Design Concepts

I created 3 designs in line with our guidelines. Together, D. and I implemented the concepts into high-fidelity mockups of the different Homegenius instances.

(click to enlarge)

(click to enlarge)

Challenge

One challenge that arose surrounded us deciding how we wanted to display the rating scale increments. In order to stay consistent with the Fannie Mae rubric, we made our scale 6-1 (6 being bad, 1 being good). Despite this not being the most intuitive scale, we thought consistency with the rubric was key, and made sure to develop a user resource key.

(click to enlarge)

Comparison Study

Comparison Study

Our next step was to create 2 comparison studies using usertesting.com. The objective of the first study was to find out what design concept 5 consumers preferred. The second study focused on 5 participants who were all real estate agents. We tested 10 people in total.

(click to enlarge)

We watched all of our testing sessions and took detailed notes on the participants as they went through the tasks we set up for them.

(click to enlarge)

Findings

Alongside finding that users preferred Option D as their overall favorite concept, We gained valuable insight while reviewing why users selected the options they chose.

Takeaways & Next Steps

Next Steps

The next step in this project is to develop more design concepts to test. These new concepts will be informed by our findings from our comparison studies. The following are guidelines we will follow closely:

  1. We will keep accessibility in the front of our minds to ensure users who are visually impaired are accommodated for as best we can. This will likely look like incorporating larger typefaces and higher contrast.
  2. All of our new concepts will incorporate a short term or sentence that will justify the score given.

What I learned

During this project, I learned what it meant to work as a part of a UX. This was my first time working with another designer. D. was awesome. She was super organized and played a key role in the competitive research, note taking and mockup designs. Likewise, I learned the importance of delegating, scheduling, and managing my time wisely while working on several projects at once. 

I was also reminded of the importance of accessibility. We were fortunate to get a participant in our comparison study that mentioned the struggles he had while trying to read the screen. Amidst being concerned with making the most visually appealing and intuitive design, I honestly left out accessibility. I’m determined to never put it on the back-burner again.

Contact me:

emmanueliheke2@gmail.com