Voice of the Customer
At the beginning of 2024, after repositioning our business, we recognized the need to optimize and streamline both our quantitative and qualitative research processes. Our goal was to establish Trela as the primary marketplace for our customers by understanding their needs in-depth and delivering a better experience.
To achieve this, we(myself and the CTO) classified our entire user base into TIERs, based on the behavior we wanted to encourage. This segmentation allowed us to prioritize conversations with High Value Customers (HVCs) and Potential High Value Customers (PHVCs)—new users who demonstrated strong potential in their first purchases.
We consolidated all research findings—NPS scores and qualitative nuggets—into a single dashboard called Voice of the Customer. This dashboard, updated daily with NPS data and weekly with qualitative research, became the central hub for distributing insights across the organization.
Business Impact
The insights from this process have driven improvements across the company:
Product Assortment: Prioritizing the most desired items based on user feedback.
Operational Efficiency: Addressing pain points in delivery and logistics.
Usability Enhancements: Refining the user experience on our platforms.
Feature Development: Identifying and implementing features that align with customer needs.
Customer Classification
We developed a set of automated rules to classify our users into TIERs within Metabase. This segmentation enabled more precise recruitment for research and informed strategic decision-making.
Objective: Focus on HVCs and PHVCs to understand their needs and identify opportunities to make Trela their primary marketplace.
Outcome: The classification became a company-wide metric, helping us set targets and track user distribution.
Automating Research with AI
Given the limited size of our design team – only me in the consumer team – we leveraged AI to streamline our qualitative research workflow. This included:
Automated Transcripts: Using AI to generate detailed transcripts from interviews.
Nugget Generation: Breaking down transcripts into small pieces (nuggets) using GPT agents.
Integration with Metabase: Automating the flow of insights from Google Sheets into Metabase dashboards using Google Scripts.
These automations eliminated time-consuming manual tasks, allowing us to focus on analysis and strategic decision-making.
Reflections and Learnings
This project reinforced the importance of streamlining research operations in a small team. By automating repetitive tasks, we created bandwidth for deeper analysis and strategy. The Voice of the Customer dashboard became a critical tool, ensuring that insights were not only accessible but also actionable.