Loan officers nowadays use a variety of outdated, old, complex LOS (Loan Origination Software) to close mortgages. Streamloan seeks to reduce the closing window by consolidating the entire process into a secure, streamlined, 100% digital and mobile solution.
Today, the average home loan for purchase takes 40+ days to close, with lender manufacturing costs upwards of $8,800. StreamLoan can reduce the closing window by consolidating the entire process into a secure, consistent, simple, private-labeled digital and mobile solution.
Input and manage agency credentials all in one place.
View a client's debts and assets at a glance.
Quit printing. Jot down notes online to share with the client.
Credit to Stubborn Library for these awesome illustrations!
As design lead, I led efforts across all 4 phases. Here is a breakdown of exactly what I did as a lead:
- Met with client to solidify his vision in terms of project requirements and branding
- Dissected company documentation to understand the product, information architecture and terms
- Solidified an 8-week roadmap for design
- Collaborated with Pre-Approval Letter design leads to strategize research plan, devised an interview script
- Conducted User Interviews with Loan Officers / StreamLoan users
- Assigned market research for design team to learn more about mortgages, lending and real estate
- Led affinity map session to boil down insights from interviews
- Met with client to discuss insights from research and plan for design stage
- Convinced client to migrate existing design files from Sketch / Zeplin to Figma
- Mapped out task flow for Pull Credit process
- Hosted sketching session with team to collaborate & ideate on feature concepts
- Delegated different portions of task flow among team
- MVP Design System to align both teams on visuals
- Elevating sketches to wireframes to high-fi using D.S.
- Validation testing with loan officers / users to confirm or deny hypotheses created from generative research in Discover phase
- Created handoff document to deliver to client & engineering, outlining structure of Figma file + naming conventions + hifi standards + file organization
Since we operated as a team of 5 designers, I knew we had plenty of manpower to borrow from. I knew that more of my time would be project-management and leading, not pushing pixels. I was also a member of a 2nd team for the same client. Therefore, my main concerns were:
Aligning the team on deliverables and quality of work.
Acting as a liason between our client and the team.
Assigning tasks and ensuring each member's desires were being met.
Product direction and facilitating project efficiency (deadlines, roadmap).
Partnering with Pre-Approval Letter team to match speed x quality.
In order to define what a valuable credit pull experience might look like, we set aside additional time for discovery during our design process. During this period, we interviewed existing StreamLoan users that are using other LOS in conjunction with the app.
We studied through our proposal document, held 4-hour meetings with our client, took way too many unnecessary walks to Starbucks, and researched similar apps that pulled credit as well as competing loan mortgage apps.
Understanding our users was invaluable to understand our next steps. We wanted to ask about their complex workflows to design better interactions for our product. Thus, we sought as much from our end users to sympathize with the challenges they face and how we could help remove those hurdles.User research was quintessential.
Loan officers are restricted to using desktops for their jobs
LOS (LendingQB, Encompass) are 20+ years outdated
Loan officers are always on the move, and need to multitask
Each has their own process to close out loans
⌚︎ Pulling credit should be quick ( < 1 minute) and effortless
Solution needs to be easily understood from desktop to mobile if officer switches devices.
Credit pull process needs to be quick and easy within constraints.
Screens should be easily digestible with most necessary info, but still familiar to old LOS to reduce barrier to entry.
Long-standing Loan Officers | 45 yrs old
Originates 5-10 loans per month
Is proactive to manage all aspects of a loan
Has worked in many banks and has been using old LOS
Gets nervous using new systems
Upcoming Loan Officers | 26 years old
Relatively new to mortgage field
Always up-to-date on new tech and software
Has only been exposed to 1 bank's LOS (LendingQB)
Appreciates change and tackling new challenges
We did some competitive research - downloaded some apps, watched demos, asked around - and realized the current LOS systems that mortgage officers used were horrendously outdated. We're talking Windows 95-lookin' programs. I felt bad for them.
I personally studied the UX and visual design patterns for Credit Sesame, a customer credit reporting company. They provided grade letters for 5 categories of credit analyses: Payment History, Credit Usage, Credit Age, Account Mix, and Credit Inquiries. I also looked into their responsive design.
I first took my own time to organize the requirements of the project. This was important to me because I wanted to organize tasks for everyone and see who could be responsible for what portions of the flow.
I then held (multiple) meetings with everyone to make sure they understood the project requirements. We then took to the drawing board to draft up iterations of concepts for each person's screens.
Initially we honed in on the main way a user would be able to pull credit since this represented the core experience of the pull credit process. As part of my process, this began as the initial requirement to show to our client to ensure that our features were familiar to what can be found in other LOS. It also helped align us on the main actions a user will take.
Once we got our client's approval on the UX and UI of the flow, we began to flesh out the various use cases. The most important use cases were the Initial State for the different types of borrowers. It took weeks to understand the requirements for all the different use cases.
When a loan officer performs a hard pull on a client, the grace period to pull again without negatively impacting the client's credit score lasts 60 days. Near the end of this period, it is common for loan officers to refresh a client's credit report to ensure that it is still in good standing and nothing substantial has changed since the initial credit pull.
I asked the users two main tasks with a given scenario. These tasks were meant to replicate their normal, everyday tasks when working with a client while they are preparing to begin shopping for homes.
We’d like you to imagine that you are working with a borrower named Priya Singh and her husband Akhil, who is the co-borrower.
Task 1 - Joint Credit Pull
Imagine that you have received Priya’s loan application and her authorization to pull credit.
Please walk us through the process of how you would go about pulling their credit from all bureaus, and how you’d save it.
Task 2 - Refresh Credit
Priya is almost set to get her loan approved, but since some time has passed, you decide to check Priya’s (joint) credit again.
Please show how you would check Priya’s credit report a second time.
Our users were happy with the simplicity of the tasks. We were given valuable suggestions, such as automatically pulling a tri-merge report instead of from 1 bureau, that we quickly incorporated into our prototypes for future testing.
Users successfully ordered a joint credit report for Priya.
Users easily navigated to Priya’s folder to refresh her credit.
Users were able to Upgrade a Report to pull from an additional bureau.
If there are already design assets, then it may be easier to skip lo-fi stages into mid-fi iterations - especially if the team workshopped sketches together with a design critique.
Take user feedback with a grain of salt - even with a small sample size. Every user can be an outlier, so wait for more users.
Take control over usability sessions, and avoid all leading questions. Each user has different experiences using this app.
I'm open to chat about any opportunities, or even just to chat!