“AI for Mutual Value Creation” Series, Part 1

*Author: Francesco Cordaro, PhD, Chief Scientist, Mutual Value Labs*

Creating mutual value (value that benefits people, organizations, and the wider systems they inhabit) begins with a shared understanding of how purpose and meaning are expressed.

Across corporate, civic, and social domains, language is where that understanding is negotiated: how institutions define responsibility, how communities express needs, and how societies imagine change.

At Mutual Value Labs, we have designed an AI-powered approach that transforms free-text language into interpretable maps of meaning, enabling organizations to understand how people construct purpose and value at scale.

By converting unstructured language — from citizen survey responses to long-form corporate strategies — into coherent, comparable representations of meaning, the approach helps researchers and business leaders identify shared priorities, emerging risks, and opportunities for more effective action.

In doing so, it creates a bridge between qualitative insight and quantitative evidence, revealing how people and institutions define purpose, negotiate value, and build collective understanding through the stories they tell and the actions they take.

Approach

Our approach combines the interpretive strengths of Large Language Models (LLM) with statistical methods that structure complex text-data, enabling us to convert raw language into statistically robust and measurable structures of meaning through a five-step process.

Text Ingestion: Breaking down long documents or survey responses into clear, manageable units of analysis.

Extraction: Using carefully designed LLM prompts to identify key elements such as actors, actions, intentions, benefits, risks, or narrative tone.
Categorization: Organizing these elements into themes using either established taxonomies or patterns that emerge naturally from the data.
Dimensional Mapping: Applying statistical techniques to uncover the deeper dimensions that reveal how people, organizations, or texts cluster by meaning.
Interpretation: Using an LLM to describe each dimension in clear, human-readable terms, turning statistical patterns into insights that can inform decision making.

This approach bridges qualitative richness and quantitative structure by turning open text into insights that help organizations, researchers, and designers uncover the meanings behind behavior, segment by story and mindset, and tailor communications, services, and interventions across domains such as health equity, food systems, and social belonging — all without flattening the complexity of human experience.

Cross-Sector Relevance

Our AI approach can be applied across diverse contexts, from social and health research (where understanding emotions, values, and coping strategies is essential) to corporate sustainability, where it helps decode how companies articulate responsibility and value creation.

Understanding Loneliness Through Language

One recent example is our work on how individuals narrate experiences of loneliness, wellbeing, and trust. Using open-ended surveys and structured narrative extraction, the methodology captured eight narrative dimensions such as emotional tone, agent positioning, moral stance, and relationships to institutions. Thousands of nuanced responses were synthesized into categories and projected into a narrative space.

Two key latent dimensions emerged: “Emotional Orientation” (from inward-facing introspection to outward-facing aspiration) and “Emotional Groundedness” (from anxious reflection to calm contentment). This then revealed four distinct mindsets: Comfortable Solitude, Hopeful Contentment, Anxious Withdrawal, and Anxious Aspiration, offering a map of emotional strategies respondents used to navigate social connection.

Decoding Sustainability Disclosures

Another example of our AI methodology is its application to corporate sustainability disclosures. By extracting activities, stakeholders, and partnerships from company reports the method reveals how firms balance their social, environmental, and governance priorities.

In a study of 46 luxury-sector companies, the analysis produced a Mutual Value Creation Matrix positioning each organization along two dimensions: Human & Social Value vs Environmental & Systemic Value and Internal Focus vs External Focus. This view exposes how different brands structure their sustainability agendas; some prioritizing internal equity and governance, others emphasizing environmental stewardship or community wellbeing, and only a few achieving a balanced approach across people, planet, and profit.

The methodology turns fragmented ESG reporting into strategic intelligence: a reproducible, data-driven understanding of sustainability positioning.

Future Applications

This AI methodology is ready to be applied across sectors and topics — from public health to service design, food systems, and social wellbeing. Potential applications include:

Identifying investment opportunities through emerging domains of shared value and cultural resonance.
Analyzing sustainability trends to understand where real progress, and new risks, are emerging.
Studying healthy ageing to inform age-friendly design and inclusive public spaces.
Understanding financial wellbeing and the emotional narratives that shape resilience.
Revealing how identity, constraint, and culture influence diet and food choices.
Supporting work on loneliness and belonging by clarifying how people navigate isolation and connection.
Analyzing organizational culture and employee voice to expose hidden structures of hierarchy, inclusion, and resistance.

This approach reveals the underlying architectures of meaning that connect individual stories with collective systems. It shows how people, organizations, and societies articulate purpose, negotiate value, and imagine change.

Using AI to Map Meaning Across Human and Organizational Systems