Property Report
Company/ Rocket Homes
Role/ Research Lead & Experience Designer
Responsibilities
Research/ plan development, execution, analysis, and reporting.
Design/ analytics evaluation, wireframes, and defining technology requirements.
Team/ 1 Product Manager, 1 Sr. User Experience Designer
Opportunity/ From competitor research and user feedback we identified a need for off-market home listings on Rocket Homes.
Goal/ Create an experience that provides information on home equity and market conditions to drive engagement with home buying and refinancing products.
Impact/ 33% increase in conversion with mortgage and refinance products.
Discovery & Define
Iterative Design Approach
A Data-Centric Approach
Our journey towards creating an optimal user experience involved two distinct approaches. Initially, we utilized a data-centric approach, analyzing information from our databases, such as market conditions, sold home listings, and property equity, to construct wireframes for user feedback. However, after just three users, a major flaw emerged - the data complexity and duplicative points hindered usability.
Refining Through User Goals
Recognizing the importance of understanding our users' objectives, we adopted a user goal approach. Each data point on the page was matched with a corresponding user goal, leading us to identify redundant sections. Through this process we reduced the content by half. The revised designs were user tested again.
User Testing & Outcomes
Research Approach
Goal/ Gather feedback on concept designs for Property Report.
Participants/ Homeowners purchasing a new home in 6 months.
Method/ Unmoderated user tests with prototypes.
Findings
Users felt the information provided could help with their decision to buy a new home or refinance their current home.
Users were surprised by the amount of public information about their home.
Users wanted more information such as crime and school ratings.
Outcomes
The team iterated on the designs using the findings and delivered to development for launch.
Measure & Iterate
Mixed Methods
The launch of the new product was in phases with an internal pilot followed by full launch. This was done in order to test data quality prior to full launch.
Pilot Survey
Goal/ Gather feedback focusing on data accuracy.
Participants/ Internal employees across the company.
Method/ Survey
Live User Test
Goal/ Gather user feedback on live experience from external users.
Participants/ Homeowners with intent to purchase in the next 6 months.
Method/ Unmoderated user tests
Findings & Outcome
Top 3 Findings
Data quality was low with many of the live test participants unable to complete their tasks due to data being unavailable.
Some survey participants mentioned similar issues, but occurrences were small.
Many participants questioned the data quality and wanted to understand how the information was populated.
Users wanted a way to ‘claim’ their home and edit their home information to update data.
Outcome
The team investigated the data issues for several months, but at some point priorities shifted, and the project was left live without updates.
One Year Later …
During prioritization it was determined to decomission Property Report due to low engagement. Reviewing prior research I took it upon myself to dig into data quality and engagement based on survey and user testing feedback to see if there were missed opportunities.
Analytics Deep Dive
Hypothesis
I began with the hypothesis that users who have all available data will have higher engagement.
Approach
I began by investigating what analytics were available. There were some critical analytics missing, but I was able to use what was available in Google Analytics to answer two questions:
How many users did, or did not, see all available data?
Comparing the two groups, what was their engagement with the experience?
Findings
I found that only 15 - 20% of our users were seeing the full experience.
Those that saw the experience were five times more likely to engage than those who did not.
After one year of being live that would equate to a missed opportunity of exposure to over a million users.
Identifying the Problem
After identifying the scale of the problem, I met with legal, product, and technology to discuss why data would not be loading. We found that during design and development we missed critical data logic that needed to be taken into consideration:
Legal restrictions varied state to state, and some states did not allow our data to be publicly viewed.
Some homes appeared without attributes (beds, baths, sq feet), a flaw in our data provider, which also caused data not to load.
Fixing the Problem
Identifying that it was both legal and data challenges we approached it in two ways:
Our team began working with our data vendor to ensure higher quality information. Including evaluating out new vendors to increase data quality.
I created wireframes of every page option with technology requirements reflecting our legal and data limitations. This created logic for the experience to change based on available information.
Outcome
While it was a long road we had two great outcomes:
The evaluation of the data provider led to the acquisition of a new vendor with higher quality data for across the company.
We went up to an 80 - 90% load rate allowing almost all users to see the available data leading to consistent 33% increase in engagement with the experience.
