Microfeedback
Capturing real-time customer sentiment to unlock personalized, data-driven experiences across Chase’s digital ecosystem.
Overview
I drove the design strategy and execution for Chase’s Microfeedback experience—crafting a scalable system to capture customer sentiment across mobile and web. Collaborating closely with product, research, and engineering, I identified key feedback moments, designed interaction patterns that were lightweight yet contextually relevant. This work laid the foundation for more personalized, data-driven experiences across the customer journey.
The Challenge
Customers often feel overwhelmed when navigating Chase’s digital ecosystem, with multiple products, features, and messages competing for attention. Without a way to express how they're feeling in the moment, it's difficult to surface what’s resonating and what’s not. We needed a lightweight, structured way to capture real-time sentiment—so we could tailor experiences, guide customers to relevant products and services, and help them find value at their own pace.
We believed that capturing customer sentiment at key moments would enable more meaningful personalization. By understanding preferences, goals, and emotional response to insights and recommendations, we could better tailor the experience. Success would be reflected in higher engagement, stronger conversion, and increased trust in Chase’s ability to guide customers toward financial value.
Process
I was appointed as the lead design liaison and strategist to partner with the Emerging Talent team and a group of engineering graduates from the University of Michigan. Together, we worked to extend the Personalization & Insights engine by designing an AI-powered system to capture customer feedback in real time and surface more relevant, actionable insights across Chase's digital experience.
Existing state of feedback capture
During the discovery phase, we identified several gaps in the current feedback capture process that were limiting the ability to gather meaningful, real-time insights:
Randomized Survey Pop-ups: Feedback was captured through occasional, non-targeted pop-ups, resulting in low response rates and irrelevant data.
Customer-Initiated Service Calls: Customers were required to reach out on their own to provide feedback, creating a reactive and fragmented experience.
Lack of Clear Incentives: There were no clear incentives to motivate customers to engage and share their opinions, leading to inconsistent feedback participation.
Limited Contextual Channels: Feedback was gathered in isolated touchpoints, lacking the contextual relevance that would allow for actionable, personalized insights.
These limitations highlighted the need for a more seamless, proactive feedback system capable of capturing relevant insights at the right moment in the customer journey—this led us to develop the Microfeedback mechanism.
Discovery
We had multiple channels and mechanisms for collecting feedback—but none driven by a centralized, customer-centric experience.
How might we build a feedback system that delivers reliable, targeted, and personalized insights?
To shape our strategy, we ran a research exercise to map out:
What feedback was being requested
How it was being captured
When customers were expected to respond
We also conducted a competitive analysis of microfeedback systems across related products and services. One insight stood out: we lacked contextual feedback—a critical gap in understanding how customers engage with our products in real time and how we can better serve their needs.
Strategy
We re-evaluated how and when insights were being surfaced to
customers. Research and analytics revealed a key pain point: customers found the nudges disruptive—often appearing while they were trying to complete other high-priority tasks.
To address this, we developed a communication framework that defined the logic for what to surface, when, and why—ensuring that insights felt timely, relevant, and additive to the customer’s journey.
To return control to the customer, we used nudges as a low-friction entry point to test our microfeedback mechanism. By embedding feedback prompts within dismissed insights, we created a lightweight, contextual interaction that let customers quickly signal relevance—without interrupting their primary task.
The adjoining image highlights this microfeedback flow and intended behavior.
Next Steps
Next, we’ll run an A/B experiment to compare immediate vs. delayed microfeedback moments and determine which pattern resonates most with customers. Success will be measured by improvements in message comprehension, response rates, sentiment scores, and drop-off reduction—using a blend of qualitative interviews and quantitative behavioral data.
The experiment will inform a single, scalable model—co-developed with cross-functional teams—that adapts based on context and timing. Long-term, the goal is to establish Microfeedback as a flexible, reusable program that meets customers where they are while driving actionable insights for the business.
This work also surfaced a key organizational opportunity: to align efforts across teams like Payments, Lending Innovation, and Internal Tools & Operations, who were independently exploring microfeedback. A centralized, cohesive approach will help reduce fragmentation and accelerate impact.