8m read
882 words
Designing a notification system for mortgage brokers
Role
Product Designer
UX Designer
UX Researcher
Prompt Designer
Tools
Figma
Figjam
Timeline
3 weeks
Team
Rocket Pro

PROBLEM
Mortgage brokers are missing critical updates due to a high volume of unprioritized emails.
There are 58 different emails that can be sent out to a single broker. That amount of emails plus how many clients a broker can have is a major contributor as to why mortgage brokers will miss emails.

SOLUTION
Introducing portal notifications
How did we get here?
IDEATION
How might we decrease the chance of our partners missing critical updates with the portal?
Designing for the portal was the most obvious solution because prior research done on mortgage brokers shows that they spend a majority of their time on a desktop device.
Web Portal
Mortgage broker's primary platform.
Current medium of notifications.
SMS Messaging
Not currently available.
Mobile App
Not currently available.
ORGANIZING
Categorizing 58 different email templates
We split the 58 different email templates into three categories. Categorizing each notification is a necessary step because it allows them to understand which notifications are urgent at a glance.
Urgent notifications
Give actionable steps to brokers to address something that's time-sensitive.
Loan updates
Inform users of workflow changes or approvals in loan pipeline.
Rocket updates
Inform users about Pro feature updates and market-related news.
EARLY DESIGNS
Showing my early designs
Once we had a good idea of what the designs might look like, we would bring that into user testing.

DESIGN JUSTIFICATION
Bucketed View vs. Full View
When designs started, there were two trains of thought: one design where a notification overlay shows all notifications in on tab and one where the notifications are separated into their own respective tabs.

TESTING
Finding what's best for our users
Partnering with the internal research team, two of my designs were used in a research study with five mortgage brokers. The goal was to understand which of the designs best suited our user's needs.

Design 1: All notifications

Design 2: Tabbed notifications
TEST RESULTS
80% of participants preferred the tabbed notifications
Of the five tests conducted, four participants preferred the tabbed notification approach. While various reasons were given, the recurring one was a desire to see the most urgent notifications without the other notifications cluttering the display.

EDGE CASES
Not all problems can be solved with one solution
One challenge I kept running into was talking myself out of designs by fixating on edge cases. Brokers aren't on the portal 24/7 and can still miss notifications. But the goal was never to eliminate missed notifications entirely. It was to reduce them. Once I reframed the problem that way, I could move forward with confidence.
FINAL DESIGN
Tab-separated notifications approach
For three weeks, I worked alongside content designers, senior UX researchers, and senior product designers to come up with the foundation of a notification system.
NEXT STEPS
Exploring notifications further
While my internship ended before I could act on these findings, the research pointed clearly toward a few directions worth exploring. All of these next steps are based on the feedback that we gathered from the A/B testing.
Explore notification preferences.
The tabbed approach assumes that brokers want all three categories equally. In reality, a broker who is buried in work might want to mute specific notifications like Rocket Updates. We also have quotes from mortgage brokers mentioning that they don't need to read specific updates. Providing a preference panel gives brokers the ability to control their notifications.
Notification panel interactions.
The current design surfaces notifications but still requires brokers to navigate elsewhere to act on them. I would want to explore how we can provide mortgage brokers the ability to take direct action from the notification itself without leaving the panel.

