UX / Product Design
U.S. Small Business Administration
When paperwork stands between a dream and reality, where do you begin?
Four questions.
The right loan.
A veteran opening a second location, a first-generation entrepreneur launching a food business, a disaster-affected owner starting over — each one has an SBA program waiting. Four plain-language questions route them to it. No program codes. No credit pull. No wrong answers.
Role
Product Designer
Type
Explorative / Portfolio
Year
2026
Deliverable
Wireframe Board + UX Rationale
01 — The Brief
The problem is the question.
SBA's existing loan finder asks: “What type of loan do you want?” The options are program codes: 7(a), 504, Microloan, EIDL. Most applicants have never heard these terms. The result is abandoned flows, wrong applications, and loan officers fielding questions the website should have answered.
The brief: redesign the loan-discovery path so any small business owner — regardless of financial literacy — can identify the right program in under three minutes, on any device, without a phone call.
The constraint: stay within the USWDS (U.S. Web Design System) component library and SBA brand guidelines. Federal accessibility compliance (WCAG 2.1 AA) is non-negotiable. No custom components that SBA's engineering team would have to maintain.
02 — The Challenge
Users can't answer the wrong question.
01
Vocabulary gap
7(a), 504, EIDL — these are internal program codes, not user goals. Asking which one a borrower wants is like asking a patient which CPT billing code they need.
02
Dead-end branching
The existing flow shows all programs regardless of eligibility. Users who don't qualify for certain programs still spend time reading them — and don't know they're wasting it.
03
Trust cliff before results
The current form asks for SSN and credit pull before showing any options. Most users abandon at this point. The quiz defers identity until after the match.
The solution reframes every question around what the user already knows: what the money is for, where their business is today, how much they need, and when. SBA program codes only appear after the match — as a label, not a question.
03 — Approach
Map programs to goals,
not goals to programs.
A veteran opening a second location. A first-generation immigrant who launched a food business from a storage unit. A disaster-affected shop owner trying to rebuild what the flood took. Each one arrived at the same wrong question on the SBA website, abandoned the flow, and called a loan officer instead — the same phone call the quiz was designed to prevent. These three users are not edge cases. They are the core audience SBA exists to serve.
Four questions, progressive disclosure
Q1 asks what the money is for — six plain-language purposes that map exhaustively to all SBA programs. Q2 establishes business stage and revenue, silently filtering programs the user wouldn't qualify for. Q3 sets loan amount using SBA's actual breakpoint values as tick marks (so users can see where the program caps sit). Q4 is optional: special eligibility that unlocks additional programs — Veterans Advantage, Community Advantage, 8(a).
No dead ends, no opaque elimination
Each answer narrows the candidate set silently. Users never see a “you don't qualify” message mid-flow. At results, the best match appears with full rationale — and two runner-ups explain why they weren't primary. Variant B (recommended in the wireframes) leads with the next step rather than just the program name, because the actual question after “which loan?” is always “what do I do now?”
04 — Wireframe Board
The complete design board
12 frames across 4 phases: hero variants, quiz screens, results, and persona walkthroughs. Built to SBA brand standards using the SBA brand guide and U.S. Web Design System component library.
05 — What This Demonstrates
UX thinking applied
to a real policy problem.
Information architecture
Restructuring a complex eligibility matrix into a linear, question-by-question flow without losing any decision signal.
When the architecture is wrong, a veteran driving to an SBDC walks away with a printout of program codes he can't read — not the application link he needed.
Progressive disclosure
Deferring complexity (program codes, eligibility details, credit requirements) until the user has enough context to process it.
Show a first-generation entrepreneur the eligibility requirements before the match, and she closes the tab before she ever learns she qualifies.
Trust-aware design
Privacy copy, no-credit-pull positioning, and transparent data use are treated as design decisions — not legal afterthoughts.
A disaster-affected shop owner who has never interacted with a federal agency needs to trust the screen before they'll touch the form.
Constraint-led design
Every component is USWDS-native. The design works within an existing federal system rather than proposing a greenfield rebuild.
A design that circumvents the federal component library won't survive an IT review — which means it never ships, and the business owner still doesn't get an answer.
06 — Usability Testing Plan
Testing the redesign
with real small business owners.
To evaluate whether the redesigned SBA.gov loan application prototype improves usability, reduces confusion, and increases user confidence compared to the current experience — specifically for first-time small business loan applicants.
Research Goals
Can users identify the correct loan type without outside help?
Can users complete the application form without abandoning or losing progress?
Do users feel confident they know what to do at every step?
Do users understand their application status after submitting?
Hypotheses
Users presented with the Loan Finder Quiz will select the correct loan type faster than users navigating the current site.
Users will complete the application form with fewer errors when inline validation and tooltips are present.
Users will feel significantly less anxious after submission when a real-time status dashboard is available.
First-time applicants will require less external help when plain language replaces agency jargon.
Methodology
Test Type
Moderated Remote Usability Test
Format
1-on-1 via Zoom or Google Meet
Session
45–60 min per participant
Participants
5–8 small business owners
Prototype
Figma clickable prototype
Moderator
Brock McClung
Task Scenarios
Task 1 — Find Your Loan
“You own a small business and need $75,000 to hire staff and buy equipment. You've heard the SBA might be able to help. Starting from the homepage, find the loan that's right for you.”
Success metric — User reaches correct loan recommendation without assistance.
Task 2 — Prepare to Apply
“Before you start your application, find out what documents you'll need and how long this will take.”
Success metric — User locates document checklist and time estimate without prompting.
Task 3 — Complete the Application
“Begin filling out your loan application. Stop after completing the Business Information section.”
Success metric — User completes section without critical errors or abandonment.
Task 4 — Upload a Document
“Your application is asking for your business tax return but your file is too large to upload. What do you do?”
Success metric — User finds alternative upload method without help.
Task 5 — Check Your Status
“You submitted your application 3 days ago and haven't heard anything. Find out where things stand.”
Success metric — User locates status dashboard and correctly identifies current stage.
Success Criteria
Task completion rate
80% or higher across all 5 tasks
Loan quiz accuracy
90% or more select the correct loan type
Form abandonment rate
10% or lower
Post-session confidence score
Average 4 out of 5 or higher
Unprompted help-seeking
Reduced versus current SBA baseline
Deliverables After Testing
- —Usability test findings report
- —Annotated screen recordings with timestamps
- —Affinity map of user feedback themes
- —Priority fix list ranked by severity
- —Updated prototype incorporating top findings
Testing Timeline
- Day 1Finalize prototype and test script
- Days 1–2Identify and confirm 5–8 participants
- Days 2–4Conduct sessions
- Day 5Synthesize findings
- Day 6Deliver findings and recommendations
This usability testing plan was designed to validate three core hypotheses — that guided loan selection, progress-saving forms, and a real-time status dashboard would measurably reduce confusion, errors, and post-submission anxiety for first-time SBA loan applicants.