Designing a better loan experience to drive growth for small businesses and lenders

As the founding designer, I helped shape end-to-end design for a startup simplifying business funding. Working with leadership and engineering, I turned complex processes into simple, trusted experiences that drove product-led growth, brought in 30,000+ users, and strengthened business performance and credibility.

Company

Role

Senior Product Designer

Headquarters

Vancouver, Canada
Industry
Fintech

Timeline

1 year
Team
CEO, CTO, 1 marketer, 2 developers
Skills
Market research
Roadmapping
Prototyping
Product design
Process development
Cover image for Levr.ai loan matcher case study

The project

Securing funding is a critical moment for small business owners, but the process is often confusing, time-consuming, and full of financial jargon. They were left feeling lost, while lenders faced incomplete or mismatched applications, leading to missed opportunities.

Levr.ai aimed to be the platform that removed this friction: helping businesses discover the right lenders, understand their options in plain language, and apply to multiple sources through a single, streamlined application. And in turn, help lenders receive complete applications.

Key achievements

As the founding designer, I worked on all aspects of design for a true zero-to-one build of Levr.ai, shaping product strategy, driving user research, and designing core experiences from scratch. I partnered closely with leadership, developers, and marketing to launch and continuously evolve the platform and internal processes.

1. Product roadmap

Defined the product direction based on insights from user interviews and market research, aligning business goals with user needs.

2. End-to-end product

Turned a complex process into a clear, confidence-building journey, supporting user growth while setting a strong foundation for product scalability.

3. Scalable design process

Built a flexible internal design process to support rapid iteration and ensure collaboration between design and development teams.

The result

My work helped Levr.ai grow to a $14M+ valuation by guiding small businesses through funding with clarity and delivering high-quality applications that brokers and lenders actively recommended, driving product adoption and company growth.
Company valuation

$14 Million

Users

30,000+

The review

“We now work exclusively through Levr.ai”

1. Product roadmap

As a venture-backed startup, we had to be laser-focused on building the right things at the right time. Balancing business milestones with real user needs, I kicked things off by deep-diving into the small-business lending space to map our best path forward.

Laying the groundwork: Research & Discovery

Understanding the audience’s pain points

With a shoestring budget and very little time, I managed to conduct a competitor analysis and a handful of interviews with small businesses and private lenders. I uncovered the critical pain points on both sides of the lending process. This allowed me to identify key opportunities to streamline the experience for both parties. You can read more on how I do this in my article How I uncover customer insights that actually shape the product.

Key Insights from business owners, brokers and lenders
😩 Business owners were:
Emotionally exhausted from multiple rejections, unclear on the process and requirements, and had security concerns around uploading sensitive data.
😤 Brokers and lenders were:

Frustrated with incomplete and unqualified applications that wasted their time and potentially led to revenue loss.

I didn’t know if I needed to do anything more. Are they waiting for me or am I waiting for them?
Small cafe owner
I’m wasting time reviewing applications that aren’t even a good fit.
Private broker
It’s a lot of information they need upfront. And, so far they’ve all been a waste of my time.
Bridal store owner
I spend all of my time on admin following up on collecting documents. I’m constantly being ghosted.
Private lender

I’d been with them [their bank] for 20 years. I couldn’t believe they said no to giving me a loan. A complete kick in the teeth.

Small business owner

I have no clue what I can qualify for. I go on Google and there’s so much info and ads, it’s overwhelming. I give up before I even start.

Entertainment business owner
I’m scared to put my sensitive information in these online forms. There’s so many scams going around.
Small business owner
It can take anywhere from 2 weeks to 3 months for my clients to fill out an application. It comes down to having complete documents and the right information.
Private broker

A clear path forward: Determining the roadmap

Committed to a user-centered design philosophy I established a user journey that would guide the business owner through the process of exploring their options and confidently applying to lenders.

This led to three core features:
1. Onboarding & questionnaire: Simplifying the starting point.
2. Loan matchmaking: Connecting businesses to the right lenders, faster.
3. Unified application: One streamlined submission to reduce friction.

I brought these insights to leadership, and together we mapped a product direction that balanced user needs with investor expectations.

High-level recreation of the roadmap when kicking-off the project

High-level recreation of our roadmap when kicking-off the project

2. End-to-end product

The product evolved through an iterative, non-linear process. The three core features were built in tandem for an MVP in just six weeks, then continuously refined over the following year.

Establishing trust: Onboarding and questionnaire

Onboarding and the questionnaire were the first touchpoint of our product for businesses, so it needed to be simple, clear, and build trust instantly.

Ideation and wireframing

Challenge: The starting point

Levr.ai had an early sign-up and questionnaire, a dense 2-step modal with ~20 mixed-format questions. It overwhelmed users on mobile and led to poor completion rates.

Goal: Reduce friction

Drawing from popular survey tools, I reimagined the flow as a guided, one-question-per-screen experience to reduce cognitive load and make onboarding feel effortless across devices.

Goal: Build confidence

To increase trust and engagement, I incorporated clear expectations, security signals, and small moments of positive reinforcement to keep users motivated through the process.

Upgrading the questionnaire format to reduce cognitive load and improve conversion

The result: 300% increase in conversion

The redesigned onboarding balanced data collection with user comfort. A step-by-step flow, clear progress indicators, and trust signals made the process simple and approachable, with gamification elements that encouraged users without feeling patronizing.

Immediate impact

✅ Completion rates jumped from ~30% to over 90%
✅ Time to complete dropped from ~16 minutes to just 2.5 minutes

Mockup of iphone with a welcome screen
Mockup of iphone with a welcome screen

Building confidence: Loan matchmaking

I started by asking: What would make reviewing loan matches feel clear, helpful, and even motivating? Drawing from user interviews, I brainstormed ways to reduce the typical overwhelm of comparing financial products and help business feel confident in their decisions. 

Design challenges and ideation

Challenge: How to best display matches

One of the first ideas was a card-style interaction inspired by dating apps; familiar, lightweight, and easy to navigate. I explored how swiping could simplify a complex process and make it feel more approachable.

Goal: Support better decisions
I hypothesized users needed two tiers of information: a quick, overall view of the loan details, and deeper details for comparing terms like rates and timelines. I also considered adding third-party ratings and reviews to build trust, giving users the external validation they rely on for high-stakes decisions.
Mockup of matches ideation on iphone
Loan matches prototype in card-swiping format

Testing and validation

Identifying the key selling points​

Quick user tests showed the most critical data points were: approval odds, loan amount, interest rate, and term length. Users wanted immediate clarity without complexity. Their visual weight and positioning were revisited to have a greater focus.

Further reinforcements required

Testing revealed that clear, confidence-boosting signals were essential. A probability of approval scale was ideated to give users an immediate sense of where they stood, without overwhelming numbers or jargon.

Users also wanted more transparency around why they were matched. To support this, I designed a criteria checklist on each match card, linking their own answers to lender requirements. This reinforced trust in the platform and gave users clarity in their decision-making.

Out of scope, for now...
I received suggestions to implement a comparison feature in order to compare selections side-by-side. This was great feedback, as it’s a familiar feature that users had come to expect in this type of situation. Despite this being a great addition, we were short on time and resources, and as a team ultimately decided this was a ‘nice-to-have’ rather than a UX breaking flow. We pushed forward with what we had and would revisit adding this feature in future iterations.

The result

We found the final design delivered the right information at the right time. Familiar interaction patterns reduced friction, while clear visual hierarchy empowered users to make confident decisions.

Overwhelm turned to confidence

✅ Most positively mentioned feature in further testing and surveys

✅ Users reported “feeling in control” of the loan process

✅ Increase in users moving from viewing matches to starting applications

Mockup of lender matches in iphone
Close up mockup of lender matches in iphone
Mockup of lender matches in desktop view

Final design for loan matchmaking results

Reducing friction: One application, multiple lenders

We set out to solve a major user pain: having to complete multiple, nearly identical applications for each lender. The vision was a single, unified application that could be sent to all selected lenders at once. We analyzed lender forms, identified common questions, and planned a dynamic experience that adjusted based on lender selection, asking for each piece of info only once, and only when needed.

Design challenges and iteration

Challenge: Time constraints
With tight timelines and technical limitations, I collaborated closely with engineering to plan a lean, phased rollout. Designs were adapted to stay feasible, focusing on clarity and usability without overcomplicating the build.
Challenge: Invasive questions
Some lenders required highly personal or complex questions, causing users to hesitate or abandon the flow. I flagged the source of each question, giving users clarity and control over which lenders they engaged with.
Challenge: Uncommon pattern

Early testing showed users didn’t realize they were completing one application for multiple lenders. To fix this, I added a visual indicator linking their progress to individual lender applications, making the process clear and intuitive.

The result

The redesigned application flow delivered results. By reducing friction and confusion, we saw a clear increase in completed applications and multi-lender submissions, meaning more small businesses secured loan offers, faster.

The streamlined experience also drove referral revenue, as more businesses applied through our platform. Lenders benefited too, with fewer incomplete applications and less back-and-forth, strengthening both partnerships and platform credibility.

The process became not just simpler, but more effective for the business owner, the lender, and for us.

Final mockup of application in iphone view
Information and feedback callouts from application
Mockup of introduction modal to application step
Mockup of document upload section in application step
Closeup of application step

Final design for the application step

3. Scalable design process

A clear, tailored design process was key to reducing errors and keeping delivery fast in our startup environment. By working closely with the team, we refined handoff, reviews, and documentation to fit how we worked best, and in return, saving time and improving outcomes for the company.

Cross-Functional handoff and review process

Collaborating for clarity and speed
Together with the team, we built a handoff process that supported shared ownership, transparency, and faster delivery. Over time, I refined the approach based on what worked best for the team.
Bringing checklists into Jira
One key improvement was embedding design documentation as actionable checklists within Jira tickets. This gave developers clear visibility into outstanding tasks and made progress more trackable, reducing ambiguity.
Changelog-style design updates
To manage mid-implementation tweaks, we introduced a changelog approach inspired by GitHub. I documented versioned updates directly in tickets, using timestamps and visual highlights to show exactly what had changed. This minimized confusion, missed updates, and unnecessary context-switching.
Developers as users
By approaching the dev team as users of the design process, we built a system that improved velocity, cut down back-and-forth, and fostered a more collaborative, product-focused culture.

Retrospective

Designing a zero-to-one fintech product taught me what building from scratch really means, shaping not just the UX, but the design process, team rituals, and product vision. I had to move fast, make clear decisions, and adapt within tight constraints.

The biggest impact came from collaboration. Working closely with developers and co-founders, I learned that great design is as much about alignment and iteration as it is about the user’s experience. I refined hand-off and communication to support engineering workflows, speeding up delivery and reducing friction.

Above all, I learned to stay focused on what matters: solutions that solve real user problems, drive growth, and scale with the business.

Image of Jennifer Daly

Jennifer Daly

Marketing @ Levr.ai, Shopify, Malwarebytes

“Rhi’s super-power was definitely her exceptional ability to balance UX/UI priorities in tandem with the evolving needs of the product roadmap. She has an incredible understanding of the customer and used that insight to develop user-centric designs that make Levr.ai beautiful software to use. Her ability to recognize customer needs and translate it into products allows the team to learn and quickly optimize.”

Related articles

Article image for "How I uncover customer insights"
How I uncover customer insights that actually shape the product
Rhiannon Mapson • 5 min read
The approach I use to run thoughtful interviews, uncover patterns, and turn insights into action, with a little help from AI.