Skip to content
TR Tye Robinson

UX case study

Mortgage Application Redesign

I led an end-to-end redesign of a digital mortgage application for first‑time buyers and existing members purchasing or refinancing. The previous experience was long, jargon‑heavy, and split between MDX and the Loan Center, causing drop‑offs, confusion, and extra work for loan officers.

The goal was to create a guided, confidence‑building journey that routes members to the right path (pre‑qual, pre‑approval, or full application), reduces friction at entry, and adds in‑app support so members can move forward without calling for help.

Role

Lead Product Designer

Company

USAA

Timeframe

Jan–Jun 2025

Team

PM, 4 Eng, QA, Compliance

Impact focus

  • Increase completion by reducing confusion and redundant steps.
  • Lower avoidable loan‑officer calls with in‑app tools and just‑in‑time guidance.
  • Improve re‑entry and status clarity for members returning later.

Audio walkthrough

Overview

Full case study

We conducted a deep dive into friction across both the online Mortgage Application and the Loan Center, with a clear goal: help members move through the process faster, with fewer errors, and more confidence at every step. Our mixed‑methods study combined a member survey (n ≈ 100), six 60‑minute loan officer interviews, competitive analysis, and end‑to‑end product walkthroughs.

Across these inputs, we heard consistent themes: confusing language around key milestones (like pre‑qualification vs. pre‑approval), a step flow that made it easy to get lost or start in the wrong place, limited in‑app tools to answer questions in the moment, and weak status and re‑entry cues when members returned to finish an application. These insights gave us a clear roadmap for change.

Problem

  • Bank jargon and unclear CTAs increased cognitive load for first‑time buyers.
  • Redundant questions and scattered groupings created context switching.
  • Lack of calculators and practical tools led to avoidable calls to loan officers.
  • Weak re‑entry and status visibility caused confusion after breaks.

Goals

  • Simplify language and instructions so members proceed with confidence.
  • Reduce redundant steps and reorganize questions into an intuitive flow.
  • Introduce in‑app tools (budget, calculators, agent connect) to enable self‑serve.
  • Make re‑entry obvious and surface status to reduce frustration and support load.

Constraints

Mortgage workflows come with real limits: regulation, required disclosures, legacy integrations, and fixed delivery timelines.

Regulatory & legal compliance +

All changes had to remain compliant with lending regulations and pass legal/risk review, which limited how radically we could simplify some language and disclosures.

Legacy systems & required fields +

The front end needed to integrate with existing back‑end services and required fields, so we focused on restructuring and clarifying the flow rather than rebuilding it from scratch or removing system‑required steps.

Design system alignment +

The UI had to align with the existing enterprise design system, which constrained typography, components, and interaction patterns and influenced how far we could push new UI ideas.

Timeline & engineering capacity +

We were working against a fixed release timeline and limited engineering bandwidth, so we prioritized high‑impact UX improvements that could ship using existing services and components.

Research plan & methods

Survey

Member survey (n ≈ 100) to understand where people got stuck, what they expected, and what pushed them to call.

Loan officer interviews

6 × 60‑minute interviews to capture recurring member confusion, rework, and the real “why” behind calls.

Evaluation

Heuristic audit + competitive review + end‑to‑end walkthroughs to identify friction and missed guidance.

Research and discovery artifacts used to align the team on pain points and priorities.

Key findings

Keep it simple (language)

Replacing bank jargon with common language improves comprehension and reduces errors.

“We need more common language for the member.” — Loan officer
  • Simplified labels and helper text
  • Clear next‑step CTAs and expectations

Intuitive flow

Prioritize key questions and group related inputs (military, assets, liabilities) to reduce context switching.

  • Remove redundancies; collapse rarely used branches
  • Use progressive disclosure instead of long pages

In‑app tools & calculators

Members need practical tools to self‑serve: budgeting, payment, taxes, and agent connection. This reduces avoidable calls.

  • Collect and connect real‑estate agent info
  • Surface calculators contextually, right when members need them

Re‑entry & status

Make return paths obvious and show what’s left. Proactive status updates reduce confusion on uploads and review steps.

  • Persistent progress and “what’s next” guidance
  • Notifications for requests, approvals, and missing documents

Recommendations → design directions

Language system

Rewrite jargon, strengthen microcopy, add examples inline.

Question architecture

Prioritize and group inputs to match member mental models.

Practical tools

Budget and payment calculators; agent connect workflow.

Re‑entry & status

Findable re‑entry, clear status, proactive communication.

Mortgage application IA

Design highlights

A few screens from the guided flow and financial information capture.

Task entry screen

Task entry

Members pick a goal (purchase, refinance) and get set‑up guidance before starting the guided pre‑qualification flow.

View prototype
Choose your next step screen

Choose your next step

A scannable hub routes members into the right flow (pre‑qual, application, refinance) with “Best for…” and clear CTAs.

View prototype
Gather income information screen

Gather income information

Step 3 of 4 (Finances) breaks income capture into clear sections and prevents mis‑categorization with targeted guidance.

View prototype

Expected impact

  • Higher completion rate via clearer language and a better‑sequenced flow.
  • Fewer loan‑officer calls due to calculators and in‑flow guidance.
  • Less confusion during restarts with stronger progress and status visibility.

Next steps

  • Prototype problem–solution pairs and run usability tests.
  • A/B test microcopy and entry routing in MDX.
  • Define event instrumentation for progress, re‑entry, and drop‑off.