Refund Recommendations

Helping members wisely spend their tax refund


Approximately 50% of our Members who deposited their tax refund into a CK Money account spent the entire amount within 45 days, often in unwise ways.

Post launching the MVP we observed a 39.1% unique CTR. Prompting us to believe that our Members were looking for such an experience. Our Members were seeking ways to help them spend their tax refund wisely.


Product Management | Engineering | Analytics | Product Legal


Product Designer


Credit Karma

The Challenge

Our objective for the next phase was to increase engagement after tax filing and boosting the 39.1 % unique CTR by 16%.


Letting our Members tell us what’s important to them; showcasing what we already know about them; and providing more financial education around what they could do with their tax refund will increase Refund Recommendation’s engagement and CTR by 16%.

Ideation Workshop

Cross-functional team leads; 10 participants

Since launching the MVP, I was unsure of how to iterate on what was shipped. We started to gather some quantitative data, but not enough to make a strategic decisions. I gathered leads from the cross-functional team for an ideation workshop. Hoping that the workshop would spark ideas for our next steps.

Ideation Workshop

Setting the Tone

The secret to expand your creative thinking is getting into the right frame of mind and most importantly ensuring you’re in a psychologically safe environment. I achieved this by starting off the ideation session with an ice breaker. Asking the group to

“Discuss alternate ways to use a paper towel roll”

Ideation Workshop

How Might We

After providing the team with some context of how the  MVP was performing, I asked the group to generate as many ideas as possible and to lean into their imagination. To help the team to come up with their ideas, I prompted the question;

How might we “optimized” and increase “personalization” to the current experience?

Ideation Workshop


To further refine our ideas, I asked the team to distill the 36 ideas into actionable priorities. We accomplished this by assessing the potential impact and effort for idea by using FigJam’s Prioritization Matrix Widget.

Subsequently, my analysis identified 3 of ideas that stood out as the most promising. With the many closely related ideas, I then grouped these similar ideas and identified a common theme amongst them.

Generated Ideas

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

Usability Testing

8 participants; 2 market segments

With the completion of the ideation workshop, the team had aligned with conducting user research.

Limited with resources, I was left with usability testing to gain qualitative data. Testing the current Refund Recommendation I sought to understand members' tax refund spending motives, actual spending habits, and their perceptions of Refund Recommendations.

Test Results

  • Focus on what’s important to them; “I only want to see recommendations that are relevant to me.”
  • More personalization; “I want to see more of what Credit Karma knows about me.”
  • Year round engagement; “I want to learn more about what I can do with my money throughout the year.”


Recommendations were presented to them like a menu for them to choose from. Creating too much of a cognitive load and inhibiting users from taking action.


Personalizing the experience by tailoring the recommendations to the users needs would boost it’s engagement and increase its click-through rate.

How might we tailor our user’s recommendations so they can better focus on using their refund towards their financial priorities.

Design Decisions

Quick Questionnaire

To help us learn what financial priority is important to our our members, I began to design a questionnaire screen. Displaying what we know about them, that way our members could make an informed decision. This questionnaire would provide personalized recommendations based on their responses.

I explored several design ideas for the questionnaire screen. Exploring ways to bring in their financial data and how they could tell us what financial priorities are important to them at this moment.

Staying within KPL guidelines, I chose the vertical stack choice chips for this interaction. While working with Content Design to refine the copy for this screen.

Design Decisions

Updated Recommendations View

Thinking back on the member feedback from the user research, I explored several ways to display their personalize recommendations. Allowing members to focus on one recommendation at a time.

Ultimately deciding to move forward with the tabbed navigation interaction. This allowed for all the necessary content to be contained on a single screen. While allowing our members to fully focus on one recommendation at a time.

Design Decisions


These recommendations are based on an alignment between Design and Product for the MVP. Without supporting data or user research, we agreed to align on these recommendations without giving users the perception that Credit Karma is giving financial advice. To learn more about this decision, click here.

To further enhance personalization, we used an 'If-Then' approach to determine the content for the recommendations.

**Click on the image to view more recommendations.


Top Findings

Personalizing the recommendations with customer data likely to drove the increased engagement.

  • TY2023- 20% increased engagement for Credit Builder and Pay off debt
  • Pre TY2023- 16% increased engagement for Credit Builder and Pay off debt

Other recommendations such as opening a Save account or Purchasing a home generally had <20% engagement. This is likely due to the lack of customer data.

Final Thoughts

What Went Well

For several months Product had been reluctant to conduct user research for any projects. But with the success with of previous user research, we had alignment with Product to move forward with the research. Ultimately, giving me the freedom to fully lead the direction of the project.

During my design process, I made sure to involve the cross-functional team every step of the way. Primarily involving Analytics, Product, Engineering, and Legal Counsel. This ensured that everyone was aware of my through process and made it easy to get buy-in and alignment from the entire team.

Top Challenges Encountered

It was a challenge with pair with Content Design for this project. Without a dedicated Content Designer for the Tax team I was forced to work with another designer that's not familiar with Tax.

During our sessions,  lot of our time was spent discussing the relevancy of this project. With Content Design questioning the merit of this project. Resulting in reducing the time we'd have to actually work on the designs.

To alleviate this, I would go into our sessions setting the context of what I am trying to achieve within our allotted time. Directing the conversation that way we don't get side tracked and start discussing other things.

Thanks for scrolling!

Discover more

Curious to know how it all began for Refund Recommendations? Explore here.