FLYE Rental
Evolving an early validated business model into a scalable, cross-channel system that enabled recurring revenue growth and operational scalability.
Design System · Brand Identity
Problem
Following its 2024 IPO, FLYE(Previous Fly E-Bike) entered a rapid expansion phase that required standardized, scalable operations. Yet its workflows remained fragmented and offline-driven, limiting growth and cross-store consistency. We needed digital transformation — but where to start and how to execute within limited organizational bandwidth?
Details
-
My role: Sole Product Designer + Product Manager
-
Duration: 2 quarters
-
Responsibilities: Execute the end-to-end design process
-
Cross-functional collaboration: C-Suite Leadership, Finance, Operations, Legal, Engineering.
Outcome
A cross-channel, multi-role service ecosystem.
Impact
Metric
Before
After
Impact
Per Rent-To-Own Order Proceessing
45min
8min
-82.2%
Per Flex Rental Order Processing
22min
5min
-77.3%
Quarterly
Finance Workload
64hrs
8hrs
-87.5%
Customer Order
Tracking
Phone Call
Real-Time App
Customer Service Workload ↓
Operations Order
Lookup
Manual Excel Search
Real-Time Dashboard
Real-Time Operational Visibility
Phase 1 - Discover
The Business Context
Post-IPO expansion made digital infrastructure a strategic priority, with Rental as the first focus.
Post-IPO Scale Pressure
Following its 2024 IPO, FLYE entered an accelerated expansion phase — requiring standardized, scalable operations across 40+ physical locations, with continued expansion into new markets.
Fragmented, Offline-Driven Operations
Without an integrated digital infrastructure, operations relied heavily on manual coordination, scattered spreadsheets, and multiple third-party tools — resulting in workflow fragmentation, operational friction, and limited scalability.
Rental Became a Core Growth Pillar
To enhance capital efficiency, rental was positioned as the foundation for scalable, asset-light growth.
The Challenges
Balancing multi-stakeholder needs within a cohesive system
Designing a shared product system that had to simultaneously serve customers, store staff, operations, finance, and leadership requires clear prioritization, explicit trade-offs, and a system that could make those decisions visible and scalable.
Designing under ambiguity and evolving requirements
Designing core product flows while business rules, pricing models, and operational policies were still evolving requiries flexible architectures, assumption-driven decisions, and designs that could adapt without constant rework.
Driving alignment in a sales-led organization
Initiating and advocating for a long-term product vision within a team historically focused on offline sales and short-term execution requires strong storytelling, cross-functional communication, and the ability to align design with business impact.
Research
Method 1: Behavioral Research through a Lean Concierge MVP
Market Signal Identified at Store Level
The rental opportunity emerged organically when several store managers identified strong demand for short-term rentals and rent-to-own options amony customers
Built a Lean Revenue-Generating Commercial Loop
Before investing in full product development, I partnered with the Operations Manager to operationalize the model using lightweight assets — including a Wix-based landing page, in-store posters, QR codes, pricing sheets, and manual payment tracking workflows — enabling real-world behavioral observation.
Gained Buy-In & Generated Design Insights
Before official app development began, three pioneer stores had already generated over $600K in profit. This live commercial execution not only aligned leadership, operations, and finance around a shared direction, but also surfaced critical insights into customer confusion, friction, and operational bottlenecks .
Method 2: Learning from Market & Infrastructure Patterns
Research Scope
-
Rental & mobility platforms
-
Rent-to-own & lease models
-
Installment-based payment systems
-
Operational infrastructure tools
Analysis Framework
-
Onboarding friction
-
Payment process
-
Contract insertion
-
Trust signals
-
Compliance design
-
Cross-channel coordination
Phase 2 - Define
Persona
Rent-To-Own Customer
Marcus Chen
Age: 29 | Delivery Courier | Low Tech Proficiency
" I can't afford a full upfront purchase but needs immediate access to a reliable bike. "
Goals
Functional:
-
Get a bike immediately without large upfront cost
-
Understand total payment and ownership timeline
-
Track payment status clearly
-
Avoid hidden fees
Emotional:
-
Feel financially in control
-
Feel supported by our services, such as free repairs
-
Feel treated like a legitimate customer, not high-risk borrower
Painpoints
-
Confusion about total payment vs weekly installment
-
Take a long time to process in-store(45min~1hr)
-
Unclear contract terms
-
Unclear payment due dates
-
Need to call store for order updates
-
Risk of high penalties if the bike is lost
Flex Rental Customer
Sofia Martinez
Age: 23 | Graduate Student | High Tech Proficiency
" E-bikes would make getting to the beach with friends this weekend much easier. "
Goals
Functional:
-
Rent quickly without paperwork friction
-
Understand pricing instantly
-
Pick-up / Drop-off easily
-
Avoid hidden fees
Emotional:
-
Feel freedom
-
Trust that the process will be easy
-
Feel modern & app-driven
-
Feel confident to start riding
Painpoints
-
Confusing rental terms
-
Don't know how to choose among many e-bikes
-
Unexpected charges, such as deposits or insurance fees
-
Manual contract signing
-
Have to wait cuase bike buyers are prioritized by store staff
Store Staff
Alan Liu
Age: 32 | FLYE Store Staff | Low Tech Proficiency
" I hope the rental process is AS SIMPLE AS POSSIBLE, since I have many retail and repair customers to handle at store. "
Goals
Functional:
-
Process each customer onboarding within 5 minutes.
-
Verify customer information and contracts quickly
-
Update bike availability, repair records and inventory quickly
-
Reduce the time for responding to rental customer inquiries
-
Avoid mistakes in payment or contract records
Emotional:
-
Feel confident handling customer questions
-
Avoid stressful peak-hour situations
-
Look professional and organized in front of customers
Painpoints
-
Preparing contracts for Rent-To-Own customers is very time-consuming
-
Searching for orders in Excel takes too long
-
Customers calling the store for order information instead of the customer service line
-
When things get busy, updates are often missed or entered incorrectly
Customer Servcie
Daniel Park
Age: 32 | FLYE Office Staff | High Tech Proficiency
" My job is ensuring that rental workflows run smoothly and issues are detected and resolved quickly. "
Goals
Functional:
-
Ensure customer onboarding processes run smoothly
-
Quickly detect problematic orders
-
Reduce manual data logging
-
Quickly access order information
Emotional:
-
Feel confident that no problematic orders are missed
-
Feel less stress when monitoring large volumes of orders
Painpoints
-
Rental data scattered across multiple spreadsheets
-
Delayed information from stores
-
Difficult to track overdue payments
-
Time-consuming manual order checks
-
Lack of centralized monitoring tools
Operations Manager
Rachel Kim
Age: 39 | FLYE Office Staff | High Tech Proficiency
" As the rental program expands, I needs tools that allow me to manage operations performance across multiple stores. "
Goals
Functional:
-
Configure and update pricing, payment cycles, and contract terms of rental plans easily without engineering support
-
Manage user permissions across roles
-
Ensure operational policies are consistently applied across stores
-
Build a collaborative system between stores to maximize the value of our store network
-
Monitor overall rental program performance
Emotional:
-
Feel confident in maintaining control over operations
-
A sense of achievement from rental revenue growth
Painpoints
-
Rental policies difficult to update (Need to explain the changes to each store manager and reprint offline materials)
-
Performance metrics require manual calculation from data scattered across documents and spreadsheets.
-
Order updates are manual and often delayed
-
Without role-based access control, everyone involved in operations has access to all materials, creating potential risks
-
Slow process when system changes require engineering involvement
Accountant
Linda Chen
Age: 41 | FLYE Office Staff | Low Tech Proficiency
" The most troublesome part is handling the tax reporting for Rent-To-Own orders and the e-bike wear cost."
Goals
Functional:
-
Quickly access accurate payment records
-
Clearly identify tax-relevant data
-
Quick access to cost-related data, such as bike and battery depreciation
-
Reduce manual reconciliation work
-
Clear visibility into each order’s details to support audit inquiries
Emotional:
-
Feel confident that financial records are accurate
-
Avoid stress during reporting and tax preparation periods
Painpoints
-
Each Rent-To-Own order requires manual tax calculations based on the reporting period
-
Depreciation data isn’t recorded and must be inferred from individual orders, have to search through multiple Excel files to calculate bike depreciation
-
Payment records spread across spreadsheets of different stores
-
Store staff can’t always ensure accurate data entry during busy hours, missed records require repeated reconciliation
The System Architecture
Designing for Scale:
The Modular Growth Strategy
Based on our business context and team bandwidth, we focused on building Rental as the structural foundation for digital transformation, while architecting the system in a modular and extensible way to support the future digitization of additional business units.
MVP Scope
1. Infrastructure Layer : The shared system foundation that powers all current and future business lines — from centralized inventory management to role-based permissions and unified user accounts.
2. Rental Module: Supports the core rental business line, including Flex Rental and Rent-to-Own.
Future Rollout
Features support other business lines: Retail, Repair, Protection Plans…
Service Blurprint
Requirement Breakdown
Phase 3 - Develop
Prototypes & Flow Mapping
Making the System Visible to Align Stakeholders
With no existing digital foundation and highly diverse stakeholder needs, documentation alone was insufficient. I used prototypes at different levels of fidelity to align stakeholders around value, constraints, and trade-offs.
Finance Teams → Accuracy & Traceability
-
Early admin platform prototypes
-
Payment calendar diagram explaining how Rent-To-Own orders are broken down into daily records to automatically calculate tax-reportable amounts.
Operations Teams → Process & Exception Handling
-
Workflow diagrams
-
Key app interfaces defining real-world workflows
CEO & Leadership → Vision & Scale
-
Leadership Dashboard Prototype(AI generated)
-
Mid-Fidelity Prototypes of All Modules(Including modules planned for future development)
Iteration
Mobile App: Iteration as Validation & Refinement before development
Prior to MVP handoff, iteration followed a structured loop of interactive prototyping, user testing, synthesis, and feasibility review.
Iteration Criteria 1
Example 1: Rent-to-Own Order Status Page
Goals:
-
Make the next required action unmistakable for users preparing for in-store pickup.
-
Encourage users to verify their identity in advance to save time.
-
Consolidate essential order details into a single, easy-to-scan view.
Iteration 1
• Order-related information was insufficient and scattered.
• 3/3 testers failed to identify the next step without guidance.
Insight: Users lacked a clear mental model of where they were in the process and what action was required.
Iteration 2
• 1/3 testers still failed to identify the next step.
• 3/3 testers ignored the verification prompt.
• Lack of a unique order identifier caused confusion when referencing the order.
Insight: Improving visibility alone was not enough—users needed explicit action-oriented guidance.
Iteration 3
• 2/3 testers still hesitated on the next step.
• 1/3 testers continued to overlook verification when it was presented as secondary content.
Insight: When multiple pieces of information competed for attention, users defaulted to status-checking rather than task completion.
Iteration 4
• 5/5 testers correctly identified the next step.
• 4/5 testers mentioned ID verification unprompted when asked what to prepare before pickup.
Outcome: Users were able to prepare for in-store pickup independently, without staff assistance—clearly understanding what to do next, where to go, and what to bring.
Example 2: Rent-to-Own Plan List Page
Goals:
-
Enable users to quickly browse available Rent-to-Own plans in their city.
-
Make pricing and plan duration transparent at a glance.
-
Support different user intents—both exploration and immediate conversion.
Iteration 1
• The plan list lacked clear visual hierarchy, making it difficult to scan and compare options.
• Users who wanted to learn more before committing were confused by a single, conversion-heavy CTA.
• Pricing details (weekly cost, deposit, taxes) were not immediately legible, increasing hesitation.
Insight: Treating all users as “ready to rent” ignored exploratory behavior and created unnecessary friction.
Iteration 2
• Users were able to distinguish plans more quickly and understand core differences at a glance.
• Exploratory users no longer felt forced into a decision and could progress at their own pace.
Insight: Supporting multiple user intentions increased clarity without sacrificing conversion readiness.
Example 3: Order Created Successfully Popup
Goals:
-
Clearly confirm that the order has been successfully created.
-
Guide users smoothly into the post-order flow without introducing hesitation or cognitive load.
Iteration 1
• 5/5 users hesitated to close the popup, fearing they might lose important information.
• The popup repeated content already shown on the order detail page, increasing cognitive load.
• Dense instructional text blurred the distinction between confirmation and next-step guidance.
Insight: A confirmation state should reassure and redirect—not require users to read or decide.
Iteration 2
• 3/3 Users felt confident closing the popup without fear of missing information.
• The transition from order creation to order management became smooth and intuitive.
• The confirmation moment reinforced system trust rather than interrupting the flow.
Insight:In post-action states, clarity and restraint are more effective than completeness.
-
Iteration Criteria 2
Example1: Rent-to-Own Order Status Page
Goal:
Provide users with all plan-related information—including pricing, payments, items, specifications, and warranty details—without increasing cognitive load or structural complexity.
Iteration 1
The page experience was clear, but the underlying structure was unnecessarily complex:
• “What’s Included” listed the items bundled with the plan.
• “Specifications” presented detailed product parameters.
• While understandable to users, the two sections overlapped conceptually and increased content fragmentation.
Insight: Separate sections improved clarity superficially, but introduced redundant data structures and additional maintenance overhead.
Iteration 2
• Users continued to access all relevant plan details without confusion or loss of clarity.
• The page became easier to maintain, update, and scale as product configurations evolved.
• Future content changes required fewer structural decisions and less coordination across systems.
Insight: Structural simplicity can often be achieved without altering perceived user experience.
Iteration Criteria 3
Example: Verification-Upload Photo ID
Goal:
Help users complete the photo ID upload step in the identity verification process, even when their situation does not match the predefined options.
Iteration 1
• Order-related information was insufficient and scattered.
• 3/3 testers failed to identify the next step without guidance.
Insight: Users lacked a clear mental model of where they were in the process and what action was required.
Iteration 2
• 1/3 testers still failed to identify the next step.
• 3/3 testers ignored the verification prompt.
• Lack of a unique order identifier caused confusion when referencing the order.
Insight: Improving visibility alone was not enough—users needed explicit action-oriented guidance.
Iteration 3
• 2/3 testers still hesitated on the next step.
• 1/3 testers continued to overlook verification when it was presented as secondary content.
Insight: When multiple pieces of information competed for attention, users defaulted to status-checking rather than task completion.
Iteration Criteria 4
Example: Verification Progress Bar
Iteration 1
• The design required multiple unique components, increasing development complexity.
• Maintaining visual consistency across different screen sizes is difficult.
Iteration 2
• The progress indicator became easier to implement and maintain.
• Development workload was reduced without sacrificing user comprehension.
• The component scaled more reliably across screen sizes and future verification flows.
Iteration Criteria 5
Example: Display of Time Format
Goal:
Present time and date information clearly and consistently across different locations and languages.
Iteration 1
• Time was displayed in a language-dependent format (e.g. “10:00 PM on Aug 11, 2025”).
• Each language required its own date and time structure.
• Localization rules varied significantly across regions, increasing implementation complexity and maintenance cost.
Iteration 2
• Time information became predictable and consistent across regions.
• Localization effort for developers was significantly reduced.
• The system became easier to extend to additional markets and future features involving time-based data.
The multidimensional table used to track the design requirements.
Internal SaaS Platform: Iterating on a Developer-Built Foundation
Development Approach
We didn't build the product strictly from design prototypes, instead, the development team directly implemented a working demo based on:
-
a existing UI component library
-
predefined data structures
-
backend operations annotated alongside the app design that affect the data on this interface.
Design Iteration
Once the working system was available, I reviewed the real workflows and interfaces and iteratively refined the design by:
-
improving layout and information hierarchy
-
simplifying operational flows
-
clarifying data visualization
-
aligning the interface with user needs
Outcome
This approach allowed design and engineering to move in parallel, enabling faster delivery while continuously improving usability.
Phase 4 - Deliver
Final Outcome
—————— Lookback ——————
Learnings
Continuous stakeholder alignment prevents costly pivots
In cross-functional teams where priorities often differ, alignment can’t be taken for granted. Regular check-ins with shared artifacts—like prototypes, flows, and visual mockups—help surface trade-offs early, keep teams aligned, and prevent late-stage redesigns that can slow things down.
Visual artifacts drive stronger buy-in than text alone
In cross-functional teams, one key role of designers is aligning people through tangible visuals. Wireframes, prototypes, and mockups make abstract ideas concrete and help bridge the gap between design work and business goals.
UX writing is tightly coupled with design
Small changes in wording can significantly change how users interpret actions and flows, sometimes leading to entirely different design solutions.
Strategic focus > perfectionism
When resources are limited, prioritizing design decisions while maintaining overall coherence is essential—otherwise it’s easy to get lost in details.
System-level thinking is essential in scalable products
In a multi-module system, no change exists in isolation. Even small adjustments required structured impact checks across related pages, flows, and touchpoints. Treating the product as an interconnected system ensured that updates strengthened overall coherence rather than introducing inconsistencies or fragmentation.
What I Would Do Differently
Build the design system alongside the first page in Figma
I didn’t formalize the design system until after the first lo-fi prototype, which led to inconsistent components and extra time spent standardizing them later. For a function-driven system like this, building the design system alongside early prototyping would have improved both efficiency and consistency—and doesn’t limit creativity.
Advocating for adequate resources early
This project required me to take on responsibilities as a design lead role. But I underestimated the scope and took on the work as it came. In addition to leading the rental system design, I was also handling marketing, brand, and other design tasks across the company. I had to work overtime every day to keep up with the workload.
Looking back, I would have communicated the project scope earlier and advocated for additional design resources. This experience taught me that design leadership is not only about delivering work, but also about setting realistic expectations and ensuring the team has the support needed to succeed.
A casual game that brings Social-Emotional Learning (SEL) practices to life through playful interaction.
Game · UX/UI · Thesis project @Parsons School of Design
