Dunbar Network Sale Qorus-NTT DATA Innovation in Insurance Awards 2026

1 .Concept and Objectives

Dunbar Network Sale represents a fundamental shift in insurance growth strategy. Traditional social referral models face a "high exposure, low conversion" problem. Our project introduces a new approach: replacing traffic breadth with trust depth.

Based on Dunbar's theory of human social limits, we use AI and social graph technology to rebuild the insurance growth engine whilst respecting human cognitive boundaries. The core innovation is moving from quantity-driven viral marketing to value-driven trust network operations, focusing on 150-person stable relationship networks rather than chasing unlimited reach.

2 .Reasons Behind

Traditional referral models face serious challenges: wasted traffic (conversion rates as low as 8-12%), high acquisition costs, trust dilution from stranger recommendations, and unsustainable short-term harvesting. When users are asked to recommend insurance to 500+ weak connections, conversion falls below 10%, wasting significant resources.

Dunbar's research shows that human stable social relationships form concentric networks with cognitive limits. Reaching 150 people requires different trust thresholds across different networks. This means insurance referrals must operate in layers, not through one-size-fits-all mass messaging.

3 .State of Competition

Current market players mainly follow the "unlimited viral" model, focusing on maximising single-touch reach. This creates high exposure but suffers from:

• Conversion rates of only 8-12%

• Customer retention rates of 40-50%

• High acquisition costs with low returns

• Trust dilution from non-contextual recommendations

Competing solutions fail to address the fundamental psychological constraint—human cognitive limits—and treat all social connections equally, missing the opportunity to leverage trust differences across relationship layers.

4 .Sources of Inspiration

The project draws its foundation from evolutionary psychologist Robin Dunbar's research on primate brain size and social network capacity. Dunbar discovered that the human brain has a cognitive limit of approximately 150 stable social relationships, organised in concentric networks:

• Core layer: 3-5 people (strongest trust, over 90%)

• Intimate layer: about 15 people (75% trust)

• Acquaintance layer: about 50 people (50% trust)

• Outer layer: 150 people (under 30% trust)

This biological constraint inspired the innovation to respect cognitive boundaries whilst using technology to amplify trust reach.

5 .Departments Involved

The project requires cross-functional collaboration across multiple departments:

• Digital Innovation Division: AI algorithm development, social graph architecture

• Product Development: Network-based referral mechanism design

• Risk Management: Social data privacy compliance framework

• Operations: Automated claims processing and fulfilment

• Marketing: Network-specific messaging and incentive design

• Data Analytics: Customer behaviour modelling and lifecycle prediction

• Customer Experience: Service optimisation and satisfaction tracking

6 .Main Results So Far

(1) .Quantitative Achievements:

• Conversion Rate: Increased from 8-12% (industry average) to 30-38% (over 300% improvement)

• Customer Retention: Improved from 40-50% to 92% (130% increase)

• Customer Lifetime Value (LTV): Grew by 50% compared to baseline

• Customer Satisfaction: Increased by 28 percentage points

(2) .Qualitative Success:

• Paradigm shift from "viral exposure" to "trust network operations"

• Successful integration of social graph technology with insurance referral systems

• Established compliance framework for social data privacy

• Validated through real-world case studies (family insurance, corporate group plans, community referrals)