Budget Direct * Policy Manager

Problem

Our customer service team was overwhelmed serving routine policy tasks, diverting them from actual high-value support and straining resources.

Solution

Revamp Policy Manager dashboard to prominently showcase renewal, claims, and cancellation features; making self-service impossible to miss.

My role

UI/UX Designer

Goal

Cut customer service routine tasks by 20% through Policy Manager self-service.

Results

18%

fewer calls to our customer service
requesting for assistance for claims, renewal
and policy cancellation

9%

uplift in policy renewal completions
through self-service flow

How We Asked

Agent Interview

Interviewed 6 customer service agents to understand pain points which led to high support volume.

Methodology
• 30-min sharing session with customer service team
• Scripted questions + open feedback
"Customers are not aware that their policies are expiring soon. Most of them do not know that they can renew their policies on our portal."
"They are not sure what they can do with our current Policy Manager."
"I have to complete quotes and renew for customers that call in daily."

Revenue-First Prioritization

Sequence dashboard features based on service frequency and revenue impact mentioned during interview session.

Priority Order

🟢 Renewal: Low traffic, high revenue gain
🟡 Windscreen/Travel Claims: Medium traffic, neutral
🔴 Policy Cancellation: High traffic, neutral

How We Analyzed

Quantitative Analysis

Tableau Exit Rate by Page
Tableau delivered precise analysis of our Policy Manager's performance, showing critical metrics like exit rate patterns across the multiple pages. We were also able to get access to total views per page to get a rough gauge of which pages have the highest traffic.
Key Findings:
• 21% of unique visitors navigated to the Renewal and Claims journey using the Policy Manager.
• "My Policies" and "My Claims" pages have the same exit rate. Both contain the most interactive engagement elements after "Home" page.
• "Home" page has 7.45% exit rate, which is lower than others, but because it has the second highest number of exits, it represents a meaningful amount of lost traffic.

Qualitative Analysis

Hotjar Behavioral Data
Through Hotjar, we were able to get qualitative insights into clicks and user frustration that numbers alone missed, revealing unique ways that users interact with the journey. Heatmaps and session recordings showed exactly which critical information users have missed, allowing us to prioritize these elements.
Key Findings:
• Users exit the "My Policies" page even though they have an expiring car policy and is due for a renewal.
• Most users start their claims journey by clicking on "My Claims".
• Policies due for renewal were buried at the bottom if customers have multiple saved quotes.

How We Designed

1. Hypothesizing

If we make Policy Manager’s self-service features more visible and clearly differentiate them through color and hierarchy, users will better understand what actions they can take and be more likely to complete them independently🎉

2. Ideating

Policy Manager Home Page

The original Policy Manager homepage served only as a static repository for existing policies and unfinished quotes, leaving users unaware of key self-service features like policy renewal, claims submission, and self cancellation.
My Design Decision(s):
Transformed homepage into self-service hub with directional buttons instantly communicating available actions.
• Switched "My Quotes" and "My Policies" link because more users visit Policy Manager to sort their policies.

Renewal First

The old My Policies page overwhelmed users with all buttons in the same red brand color which lacked hierarchy. There was also zero policy status visibility, so customers are not aware that they can renew their policy.
My Design Decision(s):
Established clear color-coded hierarchy to possible actions for each policy.
• Added status tags (Active/ Expiring Soon/Expired) with expiring policies pushed to the top for instant policy awareness.
• Added tabs to sort policies by status and product type.

Quote Completion

The Quote component in production lacked visibility on what has to be done for customers to purchase the policy. There was no urgency for customers to complete their quotes as well, which led to quotes being forgotten and expired.
My Design Decision(s):
Added prominent visual cues "Start Date Expired. Update to Purchase." to better inform users what they have to do to purchase the policy.
• Added status tag (Need Your Attention) to create urgency and prioritize incomplete quotes.
Green "Purchase" CTA to evoke trust and forward momentum.

Bonus New Feature: Self Cancellation

Previously, customers had no self-cancellation feature, so they were forced to dial customer service for policy cancellations, creating unnecessary support workload. I had to work on this project alongside the Policy Manager revamp.
Content Decision: NEW FEATURE prioritized by Business and Sales Team to give users better policy control.

My Design Decision(s):
• Made button red and secondary = to push it down the hierarchy + critical action
• Always-visible consequences = fee preview modal before confirm, disclaimers for outstanding claims
• Extra step process = deliberate friction to prevent impulse cancellations

3. AI Assisted Prototyping

In this age of AI, we leveraged Figma Make to handle complex logic like error messages and edge cases (empty states, no expiring policies); details designers often miss under time pressure. Here's the step-by-step:
Step 1: Journey Flow Prompt
Entered the full user journey into Figma Make to generate quick iterations of the Policy Manager flow, which automatically covered edge cases and animations which was used as reference by our developers.
Step 2: Embed Core UX Ideas
We injected our human ideation (green CTA + self-service hub concept etc) directly into Figma Make, ensuring the UX output aligned with our hypothesis and ideas rather than generic designs.
Step 3: Stakeholder Validation
With multiple variants provided by Figma Make, we selected ones that align with our vision as proof of concept, quickly validating it with our stakeholders who gave us their feedback, which we would then incorporate in and take note of for the actual mock-ups.
Disclaimer: AI assists with ideas but will never replace UI/UX ownership.

4. User Testing

With the prototype in hand, we conducted testing internally and with a small group of users.

Task 1:
"What can you do here?"
Task 2: "Your car policy is expiring soon. Renew it."
Task 3: "You have gotten into a car accident. File a windscreen claim."
Task 4: "You have changed your mind about your car policy. Cancel it."

Besides assigning them tasks, we also asked them questions to understand their pain points and to scope out future quality of life improvements for the Policy Manager.

Key Findings:
• 100% awareness of all features offered on the portal through Home page
• Users successfully located the expiring policy to renew their policy
• 75% of users locate windscreen claim button through product tabs on My Policies page.

With that, our hypothesis was validated. Thank you to our Business, Sales & Marketing team + our customers for willing to do the test to provide us with more insights😊

5. Iterate + Fine Tune

With user testing success confirmed and prototype approved, we created high-fidelity mock-ups to polish the UI, fixing AI-generated issues while scaling to responsiveness and production readiness.

The final mock-ups were finally ready, and development began ✅

Home Page

"My Policies" Page

"View Policy" Page

"My Quotes" Page

"My Profile Page

Testing + Launch

UAT Testing + Post-Launch Workshop

After finalizing the high-fidelity mockups and dev handoff, I oversaw the UAT implementation, rigorously testing front-end functionality and back-end integration to ensure that all scenarios and edge cases were working properly.

Post-launch, I conducted a hands-on workshop for the customer service team, walking them through the new self-service homepage so they could confidently guide customers, reducing support dependency as hypothesized while maintaining service quality during the transition to the new Policy Manager.

Post-Launch Results

18%

fewer calls to our customer service
requesting for assistance for claims, renewal
and policy cancellation

9%

uplift in policy renewal completions
through self-service flow

Reflection

Key Takeaways

1. Human oversight trumps AI speed
Figma Make might have generated multiple variants rapidly for us to reference, but we kept refining the product with feedback and insights and was able to achieve intuitive self-service navigation and successfully reduced support dependency. Only use AI as a tool and do not let it replace you entirely.

2. Data validates design hypothesis
Data powerfully validated our design hypothesis throughout the Policy Manager redesign, enabling truly data-driven UI/UX outcomes. This proves that designers armed with quantitative evidence can iteratively refine experiences to reduce friction and boost usability.

3. Rigorous UAT testing catches overlooked edge cases
Persistent front-end and back-end testing in UAT will ensure a flawless launch, proving that relentless testing is essential to deliver a robust, production-ready product. Always keep testing, and testing no matter what.

What's Next?

1. Polish Policy Manager
Conduct targeted refinements and A/B testing to optimize user engagement and flow further.

2. Rework Renewals
In Q2 2026, we will be redesigning the renewal process for all products, for a smoother and more intuitive experience.