User Behaviour Analysis - ConveGenius

Role: Senior Associate – Growth and Impact

At ConveGenius, I led a large-scale user behaviour analysis initiative to understand how government school teachers interact with AI-driven chatbots such as Smart Attendance, School-Based Assessments, and Live Reports. While quantitative data revealed usage patterns, it could not explain teacher motivations, pain points, or systemic barriers. To address this, I designed scalable qualitative feedback systems on SwiftChat, embedding conversational, emotionally intelligent flows that collected actionable insights from over 100,000 teachers across multiple states. My role required managing multi-layered communication: assessing how teachers engaged with bots, managing in-app marketing communication, and leveraging CRM channels to surface user concerns. I also worked closely with Vidya Sameeksha Kendra dashboards (Power BI) to triangulate quantitative and qualitative data, ensuring teacher voices were translated into product improvements and policy recommendations.

Challenges

Teacher adoption of chatbots was uneven across states due to network gaps, double workload from parallel registers, non-intuitive bot design, and fragmented reporting structures. Existing dashboards captured surface-level numbers, but without understanding user experience, the product risked being seen as burdensome rather than enabling.

Progress

Framework Design

Segmented users into Non-Starters, Regular Users, and New Users to tailor insights; designed state-specific conversational question banks, contextual triggers, and AI-compatible tagging systems.

Feedback Deployment

Used CleverTap to launch in-app notification surveys, achieving 10–20% engagement and 90%+ relevant responses across states like Uttarakhand, Himachal Pradesh, J&K, and Daman & Diu.

Data Integration

Analysed multilingual raw feedback alongside Power BI dashboards from Vidya Sameeksha Kendra, linking teacher sentiment with system-level metrics.

Insight Development

Identified recurring requests—edit attendance, student leave, offline/geo-tagging, simplified UI/UX, automated reports—and escalated registry/API integration needs flagged by >30% of users.

Cross-Functional Reporting

Converted unstructured teacher feedback into state-level reports with clear suggestions for product, tech, and policy teams, ensuring educators’ voices directly informed decision-making.

Impact Delivered

The analysis demonstrated that 75–84% of teachers preferred chatbots over manual registers once usability concerns were addressed, validating the potential for large-scale adoption. Insights guided the product roadmap (offline features, simplified interfaces, geo-location fixes, reporting enhancements) and informed state policy decisions around attendance protocols and assessment workflows. By embedding scalable, low-burden feedback loops, the project established a sustainable bridge between teachers, product designers, and policymakers, transforming raw teacher inputs into actionable solutions. Beyond product improvements, it highlighted how qualitative behaviour analysis and data-driven reporting can reshape EdTech strategies to better serve educators at scale.

Download PDF