A New Lens on Segmentation: Three Perspectives

In October 2025, three research experts representing different vantage points participated in a panel discussion at AMA San Diego’s Art of Marketing Conference. Their topic: a “new-to-world” methodology called qualitative segmentation, an emerging approach that uses AI-assisted analysis of qualitative data to derive meaningful customer segments from interviews, open-ended responses, or transcripts.

Segmentation has long been central to understanding customers, defining who they are, how they behave, and what drives their choices. Traditionally, marketers have relied on two established approaches: quantitative segmentation, built from large-scale surveys, multivariate analyses, and statistical modeling; and intuitive segmentation, guided by the collective experience and intuition of marketers and strategists who draw on existing knowledge, observation, and creative insight to define audience groups.

Now, a third approach is emerging. Qualitative segmentation expands the researcher’s toolkit by introducing a way to build segmentations from rich, narrative data.

The panel brought together:

• Phil Goddard, creator of the methodology and founder of Value Architect
• Kirsty D. Nunez, quantitative and qualitative segmentation strategist and President of Q2 Insights
• Kathy Townend, Head of Customer Insights at Enlyte, representing the corporate researcher’s perspective

Together, they examined how this new approach could transform segmentation practice when applied with intention, rigor, and human oversight.

‍

Phil Goddard: The Creator’s Perspective

Phil Goddard has been a qualitative researcher for more than three decades. His curiosity about how language models might expand the boundaries of qualitative insight led him to develop what he calls an Insights Modeling Pipeline, a structured, AI-assisted process that extracts, classifies, and clusters meaning from qualitative data.

The process unfolds in four deliberate steps:

1. Extraction – identifying key ideas, emotions, and values in respondents’ stories
2. Classification – mapping those ideas to core drives (the “why”) and mental models (the “how”)
3. Clustering – grouping participants with similar psychological and linguistic “fingerprints”
4. Narrative generation – translating clusters into rich, human-centered segment stories

From this process, Phil developed three segmentation approaches:

• Core Drives, which reveal underlying motivations and emotional drivers
• Mental Models, which capture how people think about and interpret the world around them
• Topic Discourse, which reflects what people are focused on and talking about most

Each of these can be used independently to create distinct segmentations, typically resulting in three or more segments, depending on the richness of the data and the diversity of participant perspectives. Together, they provide multiple pathways to uncover and organize human meaning within qualitative data.

The result is a segmentation rooted in these three complementary lenses that reveal why people think and act as they do. The approach keeps researchers actively involved at every stage, combining computational efficiency with human interpretation and oversight.

‍

Kirsty Nunez: The Researcher’s Perspective

Having practiced both qualitative and quantitative research for more than 30 years, I have seen firsthand how each method contributes uniquely to understanding people. Quantitative segmentation remains foundational, powerful for its statistical validity, predictive capability, and scalability. At its simplest, it can be as straightforward as dividing a market by observable characteristics such as gender, age, or income. However, in practice, accepted quantitative segmentation involves much more: large-scale surveys, multivariate analyses, and clustering techniques that reveal the underlying dimensions of need, motivation, and behavior.

These statistically modeled segmentations allow marketers to project results to an entire population and to quantify the relative size and potential value of each segment. Yet their rigor comes at a cost. They are time-consuming, resource-intensive, and often inaccessible to smaller organizations or those needing rapid, exploratory insight.

Qualitative segmentation fills a critical space between creative intuition and statistical precision. When Phil introduced this approach, I initially viewed it through a quantitative lens, wondering how qualitative data could be structured for segmentation. But seeing the first application changed that perspective.

In one of our early applications of this methodology, we analyzed 27 depth interviews and surfaced three clear, intuitive segments:

• Business Development Seekers
• Knowledge and Connection Cultivators
• Expertise Contributors

Each represented distinct motivations, all derived directly from participants’ words. The process was fast, transparent, and remarkably accurate.

This said, qualitative segmentation is not a replacement for quantitative segmentation. It is a complementary tool that expands the researcher’s toolkit. Its success depends on intentional design, developing qualitative discussion guides with segmentation as an explicit goal. Questions must elicit depth, variation, and motivation; otherwise, the resulting clusters risk being superficial.

Looking ahead, hybrid approaches hold enormous promise. A qual-quant fusion could begin with qualitative segmentation to reveal emotional drivers and language patterns, then use those insights to design a quantitative study that validates segment size and prevalence. Conversely, quantitative segmentation could identify macro-segments, while qualitative segmentation deepens understanding of the people behind the numbers.
This interplay, the combination of scale and story, represents the next evolution in segmentation.

‍

Kathy Townend: The Corporate Researcher’s Perspective

From a client-side vantage point, Kathy Townend sees qualitative segmentation as a way to bridge what customers say in surveys with what they mean in conversation. In B2B environments, where sample sizes may be small but insights run deep, this approach can surface underlying motivations and early warning signals that traditional tracking surveys might miss.

Kathy emphasizes two benefits:

1. Depth at low N – Small, high-quality samples can reveal complex emotional structures that help brands anticipate shifts in loyalty or satisfaction before they show up in metrics like NPS.
2. Transparency and trust – Because this methodology is researcher-in-the-loop, it avoids the “black-box” problem of many AI tools. Analysts can review each analytic step, ensuring results remain both credible and interpretable.

For corporate researchers, qualitative segmentation offers not just a new data source but a new way to frame internal conversations about customers, culture, and brand direction.

‍

A Word of Caution: Validity, Reliability, and Longevity

As with any emerging methodology, qualitative segmentation is still in its early stages. While the initial results are compelling, researchers must approach them with curiosity and caution.

The longevity and reliability of qualitative segmentation will depend on several factors:

• Consistency of inputs – The quality of the qualitative data, the structure of the discussion guide, and the richness of responses all directly influence the segmentation outcome.
• Model variability – Because language models are stochastic, each run may produce slightly different phrasing or labels. While segment patterns tend to hold, wording and nuance can vary.
• Validation over time – As with quantitative segmentation, replication across projects and populations will be key to establishing methodological confidence.
• Evolution of technology – As AI models continue to evolve, the underlying analytical frameworks may shift, requiring periodic recalibration and transparency in how segments are derived.

At this stage, qualitative segmentation should be seen as a powerful exploratory and directional tool, not a replacement for statistically validated quantitative segmentation. Its strength lies in its ability to generate hypotheses, uncover emotional depth, and guide future quantitative measurement.

The coming years will tell how well this methodology performs longitudinally, how reliably it can be reproduced, and how it integrates into broader segmentation ecosystems. For now, its promise lies in what it reveals: new ways of understanding people through the language they use and the stories they tell.

‍

Audience Insights: What’s Next

The audience Q&A revealed the curiosity and energy surrounding this new approach:

• Scale and adaptability: Early tests suggest the method performs well across both B2B and B2C contexts, provided the data is sufficiently rich.
• Variability and validation: While AI-generated narratives can vary slightly across runs, the underlying segment structures remain stable.
• Quant-qual integration: Participants were particularly intrigued by the possibility of uniting qualitative fingerprints with quantitative measures, transforming how we quantify emotion, context, and motivation at scale.

‍

A Third Pillar in the Segmentation Toolkit

Today, marketers and researchers can draw from three complementary forms of segmentation:

1. Quantitative segmentation – rigorous, predictive, statistically reliable
2. Intuitive segmentation – collaborative, creative, and experiential
3. Qualitative segmentation – narrative-based, emotionally resonant, and accessible

Each addresses different business needs. Together, they provide a more complete understanding of customers, balancing the measurable with the meaningful. As this methodology matures, it may not supplant traditional approaches, but it will undoubtedly expand the possibilities of segmentation. It empowers organizations to move faster, work with smaller but richer datasets, and bring human stories back into strategic focus.

The future of segmentation is not about choosing between qualitative or quantitative. It is about learning to orchestrate both, using the art of storytelling and the science of measurement to illuminate the full spectrum of human behavior.

Kirsty D. Nunez is President and Chief Research Strategist at Q2 Insights, a marketing research and strategy firm based in San Diego. She has spent her career helping organizations uncover deep customer insights and translate them into strategy. Today, her work focuses on integrating AI into marketing research and positioning it as a fourth category of research that complements and enhances traditional methods. She emphasizes the partnership between advanced AI tools and uniquely human strengths such as curiosity, empathy, and strategic judgment.