- Product teams face a dilemma: deep analysis or rapid development.
- AI bridges the gap, enabling both speed and depth in qualitative insights.
- Lightrun's journey illustrates a practical framework for AI-powered qualitative research.
- From manual Excel to GPT-driven insights, the process is iterative and strategic.
In today's hyper-accelerated market, product managers often find themselves at a crossroads: either conduct deep, time-consuming qualitative analysis or move quickly with superficial, gut-instinct decisions. Chen Eperman Kor, Senior Product Manager at Lightrun, shares a groundbreaking approach to overcome this challenge, demonstrating how AI can revolutionize the way product teams understand user behavior and drive strategic growth.
Lightrun, a company known for helping developers debug live applications, faced a critical juncture. With the rapid advancements in AI, they sought to expand their product lines and reach new markets. This ambition, however, presented a significant product challenge: how to make data-driven decisions at an unprecedented pace without sacrificing depth. Eperman Kor realized that while quantitative BI data provided the 'what' of user actions, it profoundly lacked the 'why'—the underlying stories, frustrations, and aspirations of their users.
Her solution began surprisingly low-tech: a manual Excel spreadsheet. By meticulously mapping user journeys, not just within the product but across their entire engagement lifecycle (from go-to-market interactions to support tickets and internal Slack conversations), she started to uncover rich qualitative narratives. Crucially, this approach focused not only on 'champion' users but also on prospects who didn't convert, revealing a treasure trove of untapped insights. This manual, iterative process allowed her to define the desired output and template for qualitative analysis before attempting to scale.
The real transformation occurred when Eperman Kor introduced AI into the equation. Leveraging tools like Cursor for data collection from disparate sources (Gong calls, BI systems, feedback platforms) and GPT for advanced analysis, she built a multi-layered system. The key was not just data aggregation but also rigorous validation and fine-tuning of the AI agent with specific business and product goals. For instance, for a debugging product, users finding bugs is a positive signal, requiring careful contextualization for the AI to interpret correctly. This iterative validation process built trust in the AI's outputs, moving beyond raw data to deliver actionable, personalized insights for different stakeholders, from executive leadership to individual product squads.
The impact was profound. Beyond informing strategic product expansion and feature prioritization, the AI-powered system enhanced customer conversations by providing product managers with a deeper, segment-level understanding of user pain points. It also proved invaluable in identifying and engaging ideal design partners for new projects, streamlining the collaboration process. While the system requires continuous fine-tuning and maintenance to adapt to evolving business goals and data landscapes, it offers a powerful blueprint for any product manager looking to unlock the strategic potential of qualitative data at scale, moving from anecdotal understanding to systemic, AI-driven insight.
“The data tells us 'what,' but it doesn't tell us 'why.' And there's a whole world of what they tried and didn't succeed at that we won't get from the data.”
- Chen Apermann Kor, Senior Product Manager
