CASE STUDY · LENDING UX · FUNNEL OPTIMIZATION
Reducing user drop-off through analytics, experimentation, and guided onboarding
ⓘ Note on confidentiality
This case study is based on real UX and behavioral analytics work conducted within a lending and decisioning platform environment.
Due to NDA obligations, privacy regulations, financial compliance requirements, and customer data protection policies, all telemetry, metrics, interaction data, workflows, and platform visuals shown in this case study have been anonymized, simplified, generalized, and detached from any real customer, organization, or production environment.
The examples shown are reconstructed representations intended to demonstrate the UX investigation process, behavioral analysis approach, product thinking, and redesign methodology without exposing sensitive business logic, proprietary systems, or personally identifiable information.
Role:
Senior Product Designer
Lending domain
Focus:
• Funnel optimization
• Behavioral analytics
Impact:
+53% funnel completion
−35% time to complete
The original onboarding flow was presented as a long single-page form with multiple financial and verification fields shown at once.
Behavioral analytics revealed a major drop-off once users reached income-related and verification steps.
Goal: Understand where users experience friction and identify the primary causes of abandonment before redesigning the onboarding flow.
The onboarding flow was designed as a single long-form application containing personal, employment, financial, and verification information.
To identify friction points, I analyzed behavioral telemetry across onboarding sections, comparing engagement, validation activity, navigation patterns, and abandonment rates.
The analysis revealed a clear bottleneck in the Income Verification section.
• Exit rate reached 48.7%, the highest among all onboarding sections.
• Generated 8.9 validation errors per session, significantly above the rest of the flow.
• Required 11.7 field corrections per session, indicating repeated attempts to complete the form successfully.
• Back navigation increased to 12.3 actions per session, suggesting uncertainty and repeated rechecking of previously entered information.
The telemetry suggested that users were not abandoning the application due to lack of intent. Instead, they were struggling to understand, validate, and correct income-related information.
This section became the primary focus for deeper investigation and redesign.
Funnel Drop-Off Root Cause Investigation
After identifying Income Verification as the highest drop-off section, I conducted a deeper behavioral analysis to understand the root causes behind abandonment.
Comparing users who successfully completed onboarding with those who dropped off revealed several consistent patterns:
• Drop-off users spent significantly more time within the section.
• Validation errors increased dramatically.
• Field corrections were substantially higher.
• Back navigation occurred more frequently.
These signals pointed to cognitive overload, unclear validation feedback, and difficulty understanding required financial information.
The combination of these factors created friction that prevented users from progressing through the onboarding flow.
Primary Insight
Users were not abandoning the application immediately after reaching Income Verification. Instead, they spent more time within the section, repeatedly corrected information, navigated backwards, and encountered significantly more validation issues before eventually dropping off.
Key behavioral patterns
• High abandonment during verification steps
• Repeated page revisits before submission
• Long hesitation before continuing
• Validation errors discovered too late
• Support requests related to “stuck” applications
100%
Started application
82%
Reached Income verification
54%
Passed Income verification
32%
Completed application
Export used to compare behavior across onboarding sections
Root cause identified
Users were not abandoning the application because of lack of intent.
Behavioral patterns indicated confusion, delayed feedback, and excessive effort required to complete income-related tasks.
Primary Friction Points
• Too much information presented at once
• No visible sense of progress
• Delayed validation feedback
• Limited flexibility to complete the application later
• No ability to save progress and continue later
Supporting observations
• Repeated scrolling behavior
• Hesitation during verification steps
• Backward navigation before submission
Before / Legacy experience
Hypothesis Statement
Breaking the long onboarding form into guided steps could reduce cognitive load, improve progress visibility, and help users complete the application with fewer errors.
Earlier validation feedback and visible progress indicators were expected to reduce back navigation, lower abandonment during verification steps, and improve overall completion rates.
Success criteria
• Increase completion rate
• Reduce validation-related drop-off
• Reduce time to complete
• Reduce back navigation
• Improve progression between onboarding steps
The redesign was guided by four principles:
• Reduce cognitive load through progressive disclosure
• Increase transparency through visible progress indicators
• Provide feedback earlier through inline validation
• Support interrupted journeys with progress recovery
Lo-Fi Solution Concept
Guided Multi-Step Flow
Problem:
Users were required to process too much information at once, creating cognitive overload and increasing abandonment risk.
Design Principle: Reduce Cognitive Load.
• Break a complex onboarding form into manageable steps.
• Present only the information required at the current stage.
• Progressively disclose additional information as users move
through the journey.
Application Progress
Goal:
Help users understand where they are in the process, what has been completed, and what happens next.
Progress Visibility:
Step indicator and 60% progress bar communicate current position clearly.
Reduced Uncertainty:
Time estimate and reassurance note reduce perceived effort and anxiety.
Clear Next Step:
Current step is highlighted and labelled, removing ambiguity about what to do next.
Increased Completion Confidence:
Prominent CTA paired with remaining-steps preview encourages forward momentum.
Lo-Fi Solution Concept
Lo-Fi Solution Concept
Early Validation
Problem:
Validation feedback only appeared after submission.
Design Principle:
Provide Earlier Feedback
Smart Guidance
Problem:
Users struggled to understand financial terminology and required information.
Reduced Cognitive Load:
Embedding help inline at the point of need removes the requirement to recall terminology definitions from memory mid-task.
Self-Service Support:
An on-demand expandable help card lets users resolve uncertainty without leaving the screen or contacting support.
Improved Data Accuracy:
Concrete income examples (salary, bonuses, self-employment) remove ambiguity about what counts, leading to more accurate user input.
Reduced Validation Errors:
When users understand what data is expected before submitting, downstream format and range errors drop significantly.
Lo-Fi Solution Concept
Lo-Fi Solution Concept
Save Progress
Problem:
Users abandoning the application before completion.
Design Principle:
Support Flexible Completion. Allow users to pause the application and return later without losing progress. Reduce abandonment caused by interruptions or incomplete sessions.
What Happens Next
Problem:
Users were uncertain about what information would be requested later in the process.
Expectation Setting:
Showing the full step sequence upfront removes surprise and reduces cognitive uncertainty throughout the flow.
Reduced Uncertainty:
The informational note explains why income data is needed, converting anxiety into informed consent.
Improved Trust:
Transparent disclosure of upcoming data requests signals respect for user autonomy and builds process credibility.
Better Decision Confidence:
A time estimate alongside a clear CTA reduces hesitation and improves forward momentum at the decision point.
Lo-Fi Solution Concept
Lo-Fi Solution Concept
Application Summary
Problem:
Users frequently navigated backward to verify previously entered information.
Reduced Back Navigation:
Displaying all entered data in a single review screen eliminates the need to move backward through steps to verify information.
Error Prevention:
Inline Edit links per section allow targeted corrections without restarting the flow, catching mistakes before they reach submission.
Increased Confidence:
Seeing a complete summary before submission gives users the reassurance needed to commit — reducing drop-off at the final step.
Faster Completion:
A clear primary CTA paired with a secondary edit path keeps the path to completion unambiguous and reduces decision delay.