Rocket Mortgage
|
UX Research Intern
May 2025 - August 2025
Timeline: 8 weeks
Role: UX Research Lead
Tools: Qualtrics, dScout, Excel, Figma Slides, Azure DevOps
Methods: Surveys, User Interviews, Literature Review
Disclaimer: Some details about this project may be limited due to a Non-Disclosure Agreement
Problem: Mortgage brokers frequently call underwriters for routine questions that require minimal effort to answer. These calls disrupt both teams’ workflows and limit brokers’ ability to self-serve within Rocket Pro. Rocket Assist (the chatbot) lacked the automation and flexibility needed to address these common inquiries.
Goal: Identify opportunities for Rocket Assist to automate common broker inquiries and reduce unnecessary phone calls.
Outcomes:
Collected insights from 130 underwriters and 6 brokers, pinpointing the most frequent and low-effort inquiries made via phone.
Identified brokers’ AI sentiment and expectations for assistive tools.
Delivered recommendations to enhance Rocket Assist’s flexibility, automation, and usefulness across varying levels of AI comfort.
Combined with my LLM research, this research helped influence a broader strategy shift for chat-based AI products within Rocket Pro.
Reflection: This project strengthened my ability to design mixed-methods research at scale, align cross-functional stakeholders, and translate insights into strategic guidance. It also deepened my understanding of AI-assisted workflows and the importance of meeting users where they are in terms of trust and adoption.
Research Deep Dive
List of Terms
Mortgages are a complicated topic. Click through this handy list of terms to get familiar with the language you may see throughout this case study
Relevant Products
Rocket Pro
Rocket Pro is a digital tool for third-party mortgage brokers to manage and monitor loan statuses when their clients have loans being processed with Rocket Mortgages and other lenders.
They use it to track loans and to submit all the necessary documents to ensure a timely closing.
Rocket Assist
Rocket Assist is a client-facing chatbot that can be found on many of Rocket's websites to guide potential homebuyers to resources they may need.
In Rocket Pro, Rocket Assist is being implemented in the Rocket Pro Portal to help with the submission of tickets for brokers who need help with a loan being processed.
Background
The primary goal of utilizing Rocket Assist in Rocket Pro is to simplify the manual processes of brokers’ workflows. This initially took the shape of moving the forms for support tickets inside the Rocket Pro Portal using Rocket Assist for submission. After successful user tests supported this implementation, the product teams began looking to the future of the chatbot for brokers.
An opportunity that both the research and product teams identified for Rocket Assist was to explore ways to minimize broker outreach to underwriters via phone calls. We know this is an issue due to existing data gathered by an operations research team about underwriter phone calls:
1.6x
1/3rd
Call volume reduction if we eliminate calls underwriters have to redirect to other teams, status updates, and document requests
7,283
Emails that coincide with a phone call per month would also be eliminated by an alternative means of question resolution
Research Approach
Research Goal
I asked the Rocket Assist product managers about blind spots that disrupted the team's ability to make decisions about the product's roadmap. They mentioned that, even if we already had an idea of the categories that broker questions for underwriters fell into, we still didn't know exactly what questions were being asked, or how much effort it actually took the underwriters to respond to them.
Were brokers asking questions that could be answered by looking at their portals? Were they calling underwriters for more complex issues that needed active assistance? This enabled us to hone in on our research goal:
Research Goal
Identify common questions & requests that
brokers call underwriters for that are prime for
automation via Rocket Assist
Research Questions
After establishing the research goal, I spent time with the conversational AI team, other UX researchers, and CX researchers discussing the research design. I sought their perspectives as a lens for considering what data would provide the most value for integrating Rocket Assist.
We knew we wanted to identify common questions that brokers call underwriters with, but we needed to gather more insights about these questions to help the team prioritize automating them.
Underwriters understood the process in responding to the questions, and could help us identify if there are questions that are common, but do not require as much complex problem-solving to resolve (ripe for automation).
We needed to know if underwriters believed an automated response could answer questions; would brokers bypass it to speak to an underwriter anyway?
With this in mind, we finalized our research questions:
In specifics, why do brokers reach out to underwriters via phone calls?
Research Question 1
What are the most common broker inquiries that require the least effort from underwriters to solve?
Research Question 2
How receptive are underwriters and brokers to automating the response to some common inquiries?
Research Question 3
Methods
Study Design
Our study utilized a mixed-methods approach with two participant groups:
Surveys: 240 underwriters invited (quantitative & qualitative)
Interviews: 100 brokers invited (qualitative)
Both groups were offered incentives to participate
Participants
Our first thought when discussing the interest in this research was that we would talk to brokers and find out why they call underwriters. After taking a moment to digest the information and think about our research plan, we quickly decided to focus the majority of our research on underwriters' experiences answering broker phone calls.
This was largely because underwriters have a higher-level overview of the questions brokers call them for. If we were to ask a single broker why they call an underwriter, we are not capturing a trend, just the behavior of an individual.
Rather than just gathering data from brokers, we could focus on studying underwriters to identify the questions brokers ask. We could then validate this data with a smaller group of brokers.
We moved forward with this because:
Underwriters also understand the effort necessary to respond to brokers
Brokers are harder to recruit and often require greater incentives.
There are different motivations behind phone calls for underwriters and brokers.
The roles of brokers and underwriters are two sides to the same coin; what you change for one workflow will affect the other.
Thus, we decided to focus on underwriters as the primary participants, and consult a smaller group of brokers as secondary participants who could give us insights while validating the initial findings.
After identifying the value in having 2 participant groups for the study, with insight from the team, we also decided to pursue a mixed-methods approach that would maximize the impact of the research.
Underwriter Survey
Coordinating with the operations team, we sent our survey to a team of 240 underwriters. We prioritized making the survey brief and punchy to maximize completion rates, ensuring that the more open-ended qualitative questions were targeted and deliberate.
Underwriters were provided with question categories based on past operations data and an open-ended option, then asked to pick their top four most common broker question categories. We then asked them to rank these categories in terms of effort, based on time taken and depth of knowledge needed to respond. Then, we had them provide examples of the 3 most common questions they receive from brokers, and categorize them.
Broker Interviews
We tapped into our broker panel inviting around 100 individuals to schedule interviews after doing an initial analysis of the survey results. This helped us validate our findings, while directly digging into brokers' experiences of calling underwriters.
Brokers were asked questions outlining their entire experience of calling underwriters. We asked them about why they called, what guidance they were looking for, the responses they received, and more about how they feel about the process. At the end, we turned the questions towards AI focusing on if and how they currently use it, as well as their thoughts on having it as an alternative for them calling the underwriters.
Outcomes
Study Participation Totals:
Survey Results
Top 3 High-Frequency, Lower-Effort Broker Question Categories:
Loan Status
Documents
Client Finances (IPAC)
Questions from High-Frequency, Low-Effort Categories
Underwriters & Brokers Agree On The Most Common Reasons for the Calls
Note: Unusual Loan Scenarios is included because it is a higher-level theme seen in underwriters' open-ended responses and mentioned by brokers in interviews that falls into multiple categories.
AI Sentiment: Underwriters vs. Brokers
Underwriters
50% of underwriters reported being "enthusiastic" or "very enthusiastic" about allowing AI to handle the responses to some of the questions brokers call with:
"I would love it. Most of the time I get asked questions that if the brokers would just look in their portal, I could get more done."
"I get asked the same questions all the time or I have to try and solve problems that aren't even relevant to underwriting."
32% of underwriters reported being "unenthusiastic" to "very unenthusiastic":
"They'll just end up calling anyway, except now they'll be mad at us if they tried to use an AI solution and it didn't work out for them."
"If we try to use AI to solve people's problems and it gives them wrong info, that's gonna hurt business."
Brokers
All 6 interviews with brokers revealed 2 common concerns with relying on AI for their questions: having to try to problem-solve with a bot and the loss of human interaction and accountability:
"I learn a lot from those phone calls. The underwriters are knowledgeable about the guidelines. I'd hate to lose those moments to a bot."
"It's just not as flexible as a person is with more nuanced situations. It can't find the grey, it just sees in black and white"
Beyond the Scope: Differing Perspectives of Phone Calls & Tension
Both groups recognize phone calls can be necessary if a loan is not straightforward due to the buyer's finances
This is a limitation both groups see in utilizing AI to answer questions.
Underwriters get frustrated by calls from brokers about information they can access in the portal
Brokers are struggling to find information in Rocket Pro or Pathfinder
Brokers typically call underwriters as a result of frustration from a lack of clarity or urgency to keep things moving
Unclear communication and low loan status visibility result in phone calls
Recommendations
In summary, the top three priority question categories were loan status updates, documents, and client finances.
To further validate these findings, we interviewed brokers to capture their experience calling underwriters with questions, their responses regarding why they call underwriters aligned with the underwriter survey data.
We also asked both groups about their experiences with AI, particularly how they felt about AI answering broker questions. Whereas underwriters showed a majority neutral to very positive sentiment (68%), brokers gave mixed responses, with all of them expressing concerns about problem-solving, and the lack of accountability when interacting with a bot.