Empathetic CUI Design for Pet Insurance Claims

Overview

This project explores how a pet insurance claim experience can be redesigned from a transactional form into a trust-building conversational system. Also, using the new claim engine to reimburse non-complex claims immediately. This will reduce internal claim processing costs and increase customer satisfaction rates.

Business Goal

Preliminary CUI study to achieve the following goals:

• ⏱ Claim drop-off rate reduction 15%
• 📉 Average claim submission time reduction 50%.
• 📞 Claim support case reduction 35%
• 🎯 Retention rate increase 25%

Project Details

• Client: Healthy Paws Pet Insurance
• Platform: Mobile App (Mobile First)
• Role: Lead UX Designer (Initial Study Group)
• Duration: 3 Months (TBD)
• Tools: Figma, UserTesting.com

Business Context

Insurance claims are the most sensitive touchpoint in the customer lifecycle.

Observed Industry Patterns

• High first-claim submission obstacles
• Elevated support volume during claim submission
• Customer distrust around reimbursement decisions
• Operational cost from manual review & OCR correction
• One competitor offers an immediate reimbursement option from connected vets.

Internal Opportunity Hypothesis

If we can:

• Reduce perceived bureaucracy
• Shorten time-to-submission
• Immediate reimbursement for none complex claim
• Increase transparency in reimbursement logic

Then we can improve:

• Claim completion rate
• Customer retention after the first claim
• Cost reduction
• Support ticket deflection
• Customer satisfaction

This project targets the reoccurring claim experience as a strategic retention lever.

User Insight (Research Synthesis)

Through exploratory interviews and behavioral mapping, several recurring patterns emerged:

Emotional Layer

• Owners feel guilt when their pet is sick
• They fear claim rejection
• They expect insurance to “look for loopholes”
• Filing a claim feels like “asking for permission”

Cognitive Layer

• Coverage language is unclear
• Forms feel bureaucratic and cold
• Users rush through forms to finish quickly
• Receipt upload is perceived as the most annoying step

Behavioral Pattern

Users are not optimizing for efficiency.

They are optimizing for emotional relief.

This insight fundamentally reframed the design direction.

Strategic Framing

Before choosing a solution direction, three approaches were evaluated:

Option A – Improved Multi-Step Form

Pros:

• Familiar pattern
• Lower development cost

Cons:

• Still bureaucratic
• Limited emotional support
• Static logic branching

Option B – Conversational UI (Selected)

Rationale:

• Linear emotional framing aligns with distressed user state
• Dynamic branching without visual overload
• Opportunity to inject supportive tone
• Enables contextual policy explanation
• Allows automated + manual fallback paths

We selected CUI not as a novelty, but as an emotional architecture.

Design Principles

1. Emotion Before Transaction

Acknowledge the pet’s condition before asking for data.

2. Minimize Cognitive Load

Replace typing with quick-select chips and AI automation.

3. Transparent System Behavior

Always explain what the system is doing (e.g., OCR reading receipt).

4. No Dead Ends

Every failure state must offer a recovery path.

5. Risk-Aware Design

Balance empathy with policy clarity and fraud prevention.

Experience Architecture

Happy Path Flow

Goal: Claim submission under 60 seconds.

1. Greeting: Empathetic acknowledgment of the visit.
2. Pet Selection: Interactive cards reduce input effort.
3. Visit Context: Progressive disclosure via category chips.
4. Receipt Capture: AI-powered OCR with confirmation.
5. Refund Preview: Immediate estimated reimbursement transparency.
6. Confirmation: Clear next steps and timeline.

Happy Path Dialog Examples

Edge Case & Risk Strategy

Designing only for the happy path creates fragility.

This system was intentionally designed with failure tolerance in mind.

A. OCR Failure

Risk:

Trust breakdown if automation fails.

Mitigation:

• Blame-free language
• Immediate retry
• Manual override option
• Real-time scan guidance (future iteration)

B. Non-Covered Item

Risk:

Perceived rejection → churn trigger.

Mitigation:

• Explain policy clearly
• Avoid harsh rejection language
• Offer alternative value (Health Diary logging)
• Provide one-tap policy access

C. Potential Fraud Pattern (Future Consideration)

Risk:

Duplicate receipts, altered amounts.

Strategic Direction:

• Background anomaly detection (non-visible to user)
• Neutral review messaging
• Avoid accusatory tone

Trust must be preserved even during verification.

Trade-Off Analysis

This solution introduces complexity.

Increased Development Cost

Conversational engine + OCR integration

Mitigation: Due to compliance issues, we are unable to use open-source for an OCR solution and server. Must be on-premise edge solution. Scoped minimum viable product with limited branching logic

Risk of Chat Fatigue

Some users prefer direct forms

Mitigation: Exit to classic form” option in the dialog. (Add exit point to each touchpoint)

Automation Failure Risk

OCR misreads damage credibility

Mitigation: Explicit confirmation step before submission: User can check the submitted claim documents and verify information during the conversation.

Principal-level design requires acknowledging these trade-offs explicitly.

System Thinking: Beyond the UI

This design impacts multiple internal systems:

• Claims processing workflow
• OCR confidence thresholds
• Support ticket routing
• Fraud detection pipeline
• Data logging for future personalization

The CUI becomes a behavioral data engine, not just a UI layer.

Future capability:

Integrate medical history continuity into conversation logic.

Measurement Framework

To validate effectiveness, the following metrics would be tracked:

Primary Metrics

• Claim completion rate (Google Analytics or Firebase)
• Time to submission (Google Analytics or Firebase)

Secondary Metrics

• CSAT after claim submission (NPS: UserTesting or Interview)
• Support tickets related to claims (Salesforce)
• OCR retake rate (Custom tracking code need to be add to the OCR code)
• Manual override usage rate (Google Analytics or Firebase)

Long-Term Metric

• Retention after first claim (Google Analytics or Salesforce)
• Claim resubmission behavior (Google Analytics or Firebase)

Qualitative User Testing:

Traditional form vs Conversational UI (UserTesting.com)

Hypothesis:

CUI improves completion and satisfaction without increasing fraud review rate.

Reflection

This project reframed claim filing from the data submission process to a vulnerable trust moment. Designing for emotion without sacrificing operational rigor is the core challenge in insurance UX. This case reinforces my design approach to:

• System-level thinking
• Emotional architecture
• Risk-aware design
• Strategic UX beyond interface polish

What I Would Iterate Next

1. Add adaptive tone based on claim type severity
2. Real-time OCR capture guidance
3. Personal medical history continuity
4. Behavioral analytics dashboard for internal teams

Next Project →