Article by ICG member John Habershon PhD
Social listening has always promised brands deep insight into what consumers really think. But in reality we’ve been stuck with an uncomfortable choice: either use automated sentiment analysis that barely scratches the surface, or manually analyse samples so small they can’t possibly capture the full richness of what people are actually saying online. Over the past year, I’ve developed and refined an approach I call Digital Conversation Analysis (DCA) – a methodology that fundamentally changes what’s possible with social listening data. Using advanced AI tools, including Claude from Anthropic, DCA lets me analyse tens of thousands of social media conversations with the same interpretive depth I’d bring to a traditional qualitative study. Most social listening projects give you either shallow automated metrics or deep analysis of small, potentially unrepresentative samples. Neither approach really captures what consumers are saying. DCA bridges this gap by combining the scale of automated analysis with the interpretive sophistication I’d normally bring to qualitative research. This isn’t automated sentiment scoring. DCA involves my active direction throughout the analytical process. I design the inquiry, interrogate emerging patterns, test hypotheses, and exercise judgment at every stage. Claude handles the processing scale while I navigate the research questions.
Answering the ‘Why’ Questions
Traditional social listening tells you what is happening – sentiment is declining, mentions are increasing, certain topics are trending. But understanding why consumers feel a certain way, why they choose one brand over another, or why particular issues matter to them? That’s typically required direct qualitative inquiry with relatively small samples. DCA changes this. Because I can analyse conversational data at real depth, it reveals the reasoning, motivations, and contextual factors that consumers articulate in their own words – at scale.
Implications for the Research Industry
For research buyers, this represents an opportunity to extract significantly more value from social data. Questions you’d previously have answered with limited samples or superficial metrics can now be explored with both depth and breadth.
Looking Ahead
Digital Conversation Analysis represents my response to a fundamental challenge in social listening: how to analyse online conversations at scale without sacrificing the interpretive depth that makes qualitative research valuable.