Translating customer sentiment into actionable insights

While exploring Citi’s internal tools, I discovered an underused customer feedback feed that captured sentiment from iOS, Android, and web platforms.

Although the data was rich, its raw format made it difficult for teams to interpret and act on. I proposed a solution to surface this feedback more effectively, and went on to design and build a data visualization dashboard that transformed scattered insights into a centralized, actionable view of customer sentiment.

This allowed product teams to better understand user pain points and prioritize improvements across Citi’s digital experience.

My role
Creative direction | Project management | Research | Sketching | UI exploration | Prototyping | UX QA

Impact and results

The final MVP implemented key features that made it immediately valuable:

  • Multi-platform information display

  • Customizable categories for different analytical needs

  • Automatic highlighting of top 3 sentiments on load

  • On-hover tooltips for detailed information

  • A data upload tool for ongoing analysis

Reviewed comments per month

5,500

Product feedback gathered in a month

58%

Design feedback gathered in a month

9%

Technology feedback gathered in a month

33%

Weekly iPhone App Feedback

Design challenge and goals

Over the course of a month, I dove into a large volume of unstructured customer feedback from iOS, Android, and web platforms. This surfaced several key challenges:

  • The feedback was scattered across multiple platforms and difficult to organize in a meaningful way.

  • Recurring themes and sentiment were buried in raw data, making them hard to identify at a glance.

  • The lack of a clear interface made it difficult for teams to interpret and act on the insights.

  • Without a centralized view, teams struggled to prioritize fixes based on real customer needs.

My goal was to build a tool that not only aggregated feedback, but made it clear, usable, and impactful—driving meaningful improvements across Citi’s digital experiences.

Design spotlight

My design process began with extensive research into visualization methods, which I validated with stakeholders. I ultimately selected a bubble chart format that effectively represented both the type and volume of sentiment data.

During initial design iterations, I shifted from a fixed sentiment sidebar to a visualization-first approach with metrics available on demand, streamlining the interface while ensuring detailed sentiment data remains accessible. This decision responded directly to user feedback and accommodated mobile requirements, creating a more versatile design that prioritizes visual pattern recognition across all device types.

I worked closely with the development team throughout implementation, adhering to strict timelines while ensuring the final product matched my approved designs.

  • Complex data becomes actionable when properly visualized; the bubble chart format proved particularly effective for sentiment analysis.

  • Automatic highlighting of top sentiments significantly reduced analysis time.

  • Data can be sensitive—without context, feedback metrics may invite unfair comparisons between teams, leading to defensiveness instead of improvement.

  • The tool's value was validated by widespread interest from other teams who proposed features for future iterations, including additional data sources, full responsiveness, API integration, and date range selection.

Key insights