Rocket Mortgage
Our Rocket Mortgage portal allows users to get a home loan online, whether they want to buy a home or refinance their current mortgage.
Application Pit Stops
Role:
UX Writer
Writing the copy, working with UX Researcher to perform content tests, working with UX Designer to iterate in working sessions, getting feedback and sign off from Product Managers.
Context:
When a user is applying for a loan with Rocket Mortgage, they can choose to fill out an online application. Many users get all the way to approval completely online. But in certain circumstances, we stop the user before they can choose their term length, interest rate, etc. and ask them to speak with a mortgage banker.
For example, when a user is looking to buy a condo, there are some additional questions a mortgage banker can ask them more efficiently over the phone.
These pit stops already existed within Rocket Mortgage, but our goal was to redesign several of them to humanize our internal teams and personalize the page to the user so that we could see how it affected metrics at this point in the loan process.
Some details have been left out due to confidentiality.
Using tools like SessionCam, I performed a complete audit of our current pit stops, including capturing live designs, content, and how users interacted with them. The tool provided helpful data like click rates, heat maps, and pathing visualizations that showed us how users got to a pit stop. I brought this info back to the UX Designer and we began brainstorming on ways to improve.
I grabbed this screenshot of the existing Rocket Mortgage pit stop from SessionCam while I was examining the clickable zones.
Based on this data, we decided that we liked the current structure and message, but we could still take home some simple wins. For one, our style guide suggests clickable components should be written in the user’s voice. By rewriting the expandable questions in first person, we achieved a greater level of personalization and system consistency. We also added the user’s name to the page to remind them that this content was curated and served up specifically for their circumstances.
Those pieces checked off the personalization aspects, but what about ways in which we could humanize the lender? We started to look at examples of humanization within other digital experiences.
I found a couple pretty funny examples of failed personalization in an article by Rex Su. One was a chatbot given a human-like appearance, yet she was fairly emotionless. Her eyes also followed the user’s mouse around the screen for an extra dose of uncanny. Another example was an intrusive autoplay video of a human paired with a CTA of “Find your products here!” The button didn’t actually lead to any sort of interaction with a person, but merely exploited the fact that human attention is drawn to other humans. A tactic like this could increase your CTR, but would most likely also increase your users’ dissatisfaction rating.
The main takeaway here was that if your experience is entirely digital, there’s no need to force humanity where it doesn’t belong. This was great insight which reinforced our decision to include images of real people on our pit stops. There are real people involved in the mortgage process, and our users should be expecting them.
We also looked into including quotes from our bankers to accompany the images. However, we ended up not acquiring any due to time constraints. With this information at hand, we began our working sessions.
Our final decisions included:
Adding imagery of our real bankers
Using our clients name in the copy
Rewriting the FAQ in first person
Revamping other vague and unhelpful content
We ensured all these decisions were feasible with our stakeholders, including our Product Managers and the Developers. I then worked with our Research Team Leader to perform a content test on the pit stops. This test helped to identify remaining areas of confusion and places in the copy where the tone didn’t feel appropriate for the situation. I edited the content, got sign off from stakeholders once again, and we handed it off to the developers.
Lastly, the Product Managers and our Developers launched some additional metric tracking for this feature. While I can’t get into the results, this feature’s redesign inspired the team to start keeping track of at least one metric that was not previously documented and will lead to more robust reporting in the future.