Credit Karma Tax

Helping Credit Karma members make smarter financial decisions with their tax refund

Tax Refund Planning

Overview

During tax season, millions of Credit Karma members receive large tax refunds—but spend them quickly and impulsively. Internal data showed that 50% of members spent their entire refund within 45 days, often on short-term wants.

Recognizing an opportunity to guide smarter financial decisions, our team launched a basic MVP of Refund Recommendations—a feature that surfaced ways members could use their refund. After a strong 39.1% unique CTR, we knew members were interested, but the experience lacked personalization and clarity.

This case study explores how I led the redesign of Refund Recommendations—from a generic MVP to a more relevant, personalized experience—that drove increased engagement and deeper alignment with members’ financial priorities.

Role:

Product Designer — led UX strategy, research, visual design, and execution

Timeline:

8 weeks

The Challenge

The MVP presented several refund use cases at once, like a static menu. While it validated user interest, it overwhelmed members with too many options and lacked relevance.

Business Goal

Increase post-tax filing engagement by 16%, and drive adoption of Credit Karma products like Credit Builder and Save.

User Goal

Help members feel more confident about how they use their refund by surfacing personalized, relevant guidance aligned with their current financial priorities.

What's Currently Happening

The initial experience was one-size-fits-all: members were shown multiple unrelated options, regardless of their actual needs. We had early performance signals—but lacked insight into why it worked, and how to improve it.

Hypothesis

If we let users express what matters to them, surface what CK already knows about their financial situation, and provide light guidance—not advice—we can increase engagement and make the experience feel more useful and trustworthy.

Process

Cross-functional Ideation Workshop

Brought together 10 leads across Design, PM, Eng, Analytics, and Legal.
Prompt: How might we increase personalization without overwhelming users?
We generated 36 ideas and prioritized them using a matrix to find themes with high impact and low effort.

Top themes:

  • Provide Members with more personalization
  • Understand what’s important to our members
  • Provide more financial education and literacy for our Members

User Research & Usability Testing

8 participants across 2 segments. Focused on:

  • Spending behaviors after receiving refunds
  • What made a recommendation from the initial experience feel relevant

Top Insights:

  • “I only want to see what applies to me.”
  • “CK knows me—why not use that?”
  • “Help me plan for the whole year, not just tax season."

Design Decisions

Quick Questionnaire

Let members self-identify their top priorities (e.g., build savings, buy a car, pay off debt).
Used vertical stacked chips (KPL-compliant) with copy refined in partnership with Content Design.

If-Then Logic Personalization

We used member data and questionnaire inputs to conditionally render tailored content:

  • No Save account? → Prompt to open one.
  • Already using Credit Builder? → Suggest strengthening usage habits.
  • Declared “buy a car”? → Surface personalized auto loan steps.

Simplified Decision Making

Focused interface showing one recommendation at a time. I tested vertical scroll, horizontal cards, and tabbed navigation—ultimately choosing tabs for clarity and emotional focus.

Build Trust Through Feedback

While collaborating with stakeholders we decided to add feedback that would make the experience feel more trust worthy. Providing users with the feel of an "advisor" and nothing static.

Data Visualization- Used data visualization to show potential financial impact.

Loading Dots- Added loading dots to signal thoughtful processing.

Final Designs

Impact

Key Results

+20%

Credit Builder and Pay Down Debt

<20%

Recommendations without user context

What I Learned

  • Personalization doesn’t require deep AI—just relevance, clarity, and user involvement
  • Early alignment with key stakeholder helped prevent blockers
  • Cross-functional visibility during design made it easier to get buy-in and ship fast

Final Thoughts

This project transformed a generic MVP into a personalized, goal-oriented experience that empowered members during a financially critical moment. The new design helped increase clarity, reduce overload, and drive better financial decisions—while directly improving product engagement.

Thanks for scrolling!