Stories, Complaints, and Confusion: Structured Analysis of Qualitative data

Last year our session focused on why qualitative data matters. The reception was quite good, and I think we piqued everyone’s curiosity, and perhaps exploration. This year our talk addresses the next question: what does it actually mean to operationalise qualitative data so that it influences real decisions? This year we would also focus the presentation through the lens of product companies such as D2C, digital consumer applications, so that it is more relatable to the Vizchitra audience. We want to be able to speak to those who have to support decision making using data, looking at how data is analysed, and how the insights are expressed to enable swift, and sure decisionmaking.

The core idea of this talk is that qualitative data become really useful when we can look at various sources of data, and then analyse them structurally rather than only narratively. Product analytics platforms already do this with behavioural data: they construct cohorts, segment users by shared attributes, and examine patterns across variables. We argue that qualitative information can be analysed similarly — not by reducing it to numbers, but by organising it so patterns, relationships, and evidence chains become visible. To make this concrete, the talk will use a synthetic case study. We would introduce a fictional product company attempting to answer a realistic question about user satisfaction and retention. We will share a journey using combined datasets: NPS responses with written feedback alongside excerpts from research calls. We’ll use this case study to show how with these sources of data, and structuring using meta data and attributes can lead to useful insights that can drive decisions.

This topic matters to us because we think that this is a glaring blindspot for many teams. And with the swift adoption of consumer AI tools people are dumping qualitative data into chat windows, expecting deep insights when all they’re going to get are summaries.