Original Analytical Models

10+ Frameworks for Behavioural Intelligence

These models were developed as part of the portfolio methodology. Each provides a specific analytical lens for understanding how behaviour creates, destroys, or redirects value.

Behavioural Value Chain

Mapping the full path from signal to outcome

What it asks

Where does intended value fail to become realised value?

Why it matters

Most business problems are diagnosed at the symptom level. The value chain forces diagnosis at each stage where behaviour can leak, distort, or redirect value.

How it works

Trace every investigation through eight stages: Signal → Perception → Interpretation → Decision → Action → Habit → Outcome → Feedback. At each stage, identify the behavioural mechanism and potential leakage.

Example application

An AI adoption programme shows high initial usage but declining repeat use. The value chain reveals the leak occurs between Action and Habit — the behaviour does not stick because workflow redesign and manager reinforcement are absent.

Business relevance

Transformation, adoption, customer experience, risk management.

Behavioural Leakage Audit

Finding the gap between intended and realised value

What it asks

How much value is being lost between each stage of the behavioural value chain, and through what mechanism?

Why it matters

Organisations often measure inputs and outputs but not the behavioural mechanisms that connect them. The leakage audit quantifies where value disappears.

How it works

For each stage, estimate the conversion rate, identify the dominant behavioural mechanism, and calculate cumulative leakage.

Example application

A customer onboarding flow has 80% initial engagement but only 20% sustained usage. The leakage audit identifies three friction points — cognitive load at signup, unclear value at day 7, and absence of social reinforcement at day 30.

Business relevance

Customer experience, product adoption, employee engagement, operational efficiency.

Behavioural Elasticity Map

How responsive is the behaviour to intervention?

What it asks

Which behavioural mechanisms are most and least responsive to intervention, and at what cost?

Why it matters

Not all behaviours are equally changeable. Investing in high-effort interventions on inelastic behaviours wastes resources.

How it works

Score each behaviour on changeability and impact. Plot on a 2×2 to prioritise.

Example application

In a consulting firm, knowledge reuse behaviour is highly elastic (responds to search design and incentives) while risk-reporting behaviour is inelastic (driven by regulation and culture).

Business relevance

Resource allocation, intervention design, transformation prioritisation.

Behavioural Segmentation

People in the same group behave differently

What it asks

What meaningful behavioural states exist within a population, and what drives the differences?

Why it matters

Demographic or role-based segmentation often masks the behavioural variation that matters most.

How it works

Identify behavioural indicators, cluster individuals by behavioural patterns rather than demographics, and diagnose separating mechanisms.

Example application

Developers show three distinct AI tool adoption patterns — enthusiastic experimenters, pragmatic evaluators, and passive resisters. Each requires a different intervention.

Business relevance

People analytics, customer insights, product strategy, change management.

Trust Calibration

Is reliance appropriately calibrated?

What it asks

Are people trusting the right things, to the right degree, for the right reasons?

Why it matters

Both over-trust and under-trust create business risk.

How it works

Map trust objects, measure current trust levels, assess appropriate levels based on capability and reliability, identify calibration gaps.

Example application

Analysts over-trust AI-generated summaries but under-trust internal knowledge systems — leading to both error risk and duplication.

Business relevance

AI adoption, risk management, technology transformation, governance.

Behavioural Debt

The compounding cost of behavioural shortcuts

What it asks

What accumulated behavioural shortcuts, workarounds, and misaligned incentives are creating hidden costs?

Why it matters

Like technical debt, behavioural debt compounds over time. The cost becomes visible only in crises.

How it works

Audit current practices for workarounds, informal systems, incentive misalignments, and shadow processes. Estimate cumulative cost.

Example application

A consulting firm discovers 30% of analyst time is spent recreating existing knowledge — behavioural debt from years of rewarding individual output over collective reuse.

Business relevance

Organisational effectiveness, culture assessment, M&A due diligence.

Decision Archaeology

Understanding how we got here

What it asks

What historical decisions, incentive structures, and organisational norms created the current behavioural patterns?

Why it matters

Present behaviour is shaped by past decisions. Understanding the archaeology reveals why previous interventions failed.

How it works

Trace current patterns back to originating decisions, incentive designs, leadership signals, and environmental conditions.

Example application

A company's resistance to data sharing originated in a decade-old data breach — now misapplied to internal collaboration, creating knowledge silos.

Business relevance

Change management, culture transformation, post-merger integration.

Human–AI Boundary Map

Where should humans and AI divide responsibility?

What it asks

For each task and decision, should the human delegate, verify, override, or retain full accountability?

Why it matters

The question is not whether to use AI but where the boundary should sit.

How it works

Map tasks on AI capability × error consequence. Assign boundary types: delegate, verify, collaborate, or retain.

Example application

In risk assessment, AI excels at pattern detection (delegate) but humans must retain judgement on novel situations (retain).

Business relevance

AI transformation, technology governance, workflow design, responsible AI.

Evidence Confidence Score

How much should we trust what we think we know?

What it asks

What is the confidence level of each piece of evidence?

Why it matters

Not all evidence is equally reliable. Acting on low-confidence evidence as if it were high-confidence leads to overconfident interventions.

How it works

Score each source on credibility, methodology quality, recency, replicability, and relevance. Aggregate into a confidence score.

Example application

A company claim about AI adoption rates scores high on recency but low on methodology transparency.

Business relevance

Research quality, decision-making under uncertainty, evidence-based strategy.

Falsification Framework

What would change the conclusion?

What it asks

What specific evidence would disprove the current hypothesis?

Why it matters

The most important intellectual discipline is stating what would change your mind.

How it works

For each hypothesis: (1) what observation would disconfirm it, (2) what alternative explanation would be more parsimonious, (3) what experimental result would require abandoning the intervention.

Example application

If employees with high trust scores show identical adoption rates to those with low trust scores, the mechanism is not trust.

Business relevance

Research integrity, hypothesis testing, evidence-based decision-making.