Transforming User Insights into Actionable Design Insights with AI and Analytics

The transformation of our design insights process for the account management portal at Pitney Bowes has revolutionized how we handle customer feedback. By integrating AI tools like Chat GPT and Power BI, we streamlined the conversion of raw data into actionable insights. This initiative, which I spearheaded, empowered our UX team to quickly understand and address customer needs, improving satisfaction and user experience. Creating UX empathy maps was a game changer, providing a comprehensive view of customer emotions and behaviors and facilitating a user-centric design approach. Overall, this project has enhanced our ability to deliver exceptional user experiences and drive continuous innovation.

Introduction

One of my team’s focuses is enhancing the account management portal for Pitney Bowes. This platform is crucial as it enables customers to manage their relationships with us, pay bills, and seek support. Given its significance as a primary touchpoint, we must provide an exceptional user experience. Our primary KPI is the Net Satisfaction (NSAT) score, which measures customer satisfaction. Historically, we relied on Excel documents filled with customer verbatims. While insightful, it was time-consuming to distill this data into actionable steps.

I recognized this challenge and envisioned leveraging AI to streamline and enhance our data analysis process. This case study outlines our journey to transform these insights into a powerful tool that drives innovation and enhances user experience.

Data Sources and Tools

Feedback Form:
Customers log out and see a pop-up window asking how satisfied they were with their experience, followed by specific questions. This captures valuable qualitative data.

NSAT Data:
NSAT measures customer satisfaction on a 5-point scale, similar to CSAT and NPS. It’s calculated by subtracting the percentage of “Dissatisfied” and “Very Dissatisfied” responses from the “Satisfied” and “Very Satisfied” ones.

Fullstory:
Fullstory is a screen recording software that allows us to observe customer interactions with our interface. It provides visual insights into user behavior.

Chat GPT:
We use Chat GPT to categorize, summarize, and analyze customer verbatim feedback. This AI tool helps us quickly process large volumes of qualitative data.

Power BI:
Power BI visualizes and interprets data. It helps create dynamic dashboards and reports, facilitating data-driven decision-making.

Noteworthy

When I joined Pitney Bowes, our NSAT rating was 6; it’s now in the 80s for Account Managment.

Methodology

Before this initiative, we received raw, separate data that was challenging to process and act upon efficiently.

 

Data Collection:

We gathered data from feedback forms, NSAT scores, customer verbatims, and Fullstory screen recordings continuously to ensure comprehensive understanding.

Data Integration:

We integrated these data sources using Chat GPT to categorize and summarize customer verbatims, extracting key themes and sentiments. This AI-driven analysis quickly identified patterns.

Data Analysis:

Power BI helped create dynamic dashboards, merging data from feedback forms, NSAT scores, verbatims, and Fullstory recordings. This presented the data in an easily digestible format.

Creating UX Empathy Maps:

Using Chat GPT, we created UX empathy maps, organizing feedback into “say,” “think,” “do,” and “feel” categories. This provided a comprehensive view of user experiences, making our design approach more user-centric.

Implementation

Setting Up the System:

Integrating Chat GPT and Power BI into our workflow was crucial. We configured these tools to work seamlessly with our data sources.

Training and Onboarding:

I conducted comprehensive worshops sessions with my designers. We learned to interpret Power BI visualizations and use Chat GPT insights.

Challenges and Solutions:

I worked with our solution architect to refine data integration processes and provided ongoing support and training to the team, fostering ownership and acceptance.

By integrating these methodologies, we transformed raw data into actionable insights, enhancing our understanding and improving the customer experience.

Conclusion

The design team now has robust information at the start of projects. With integrated data sources and tools like Chat GPT and Power BI, they quickly understand the customer mindset. We have linked Fullstory sessions to the actual customer leaving feedback. All of this, plus an AI-generated empathy map, allows my team and I to identify and address pain points effectively, driving innovation and improving user experience.

By transforming raw data into actionable insights, we’ve empowered our UX team to make data-driven decisions that reflect customer needs and desires. This initiative has streamlined our processes and significantly improved our ability to deliver exceptional user experiences.