Agent-to-Agent Negotiation in Airline Revenue Management
Designing engagement architecture for machine-to-machine commerce at scale
The Challenge
An airline redesigned its revenue management system to accommodate a world where an increasing proportion of ticket purchases were initiated by AI agents acting on behalf of travellers. These machine customers did not browse fare calendars or respond to urgency cues ('only 3 seats left'). They queried availability programmatically, compared across carriers simultaneously, and optimised for complex multi-dimensional mandates (price, timing, loyalty status, carbon footprint, connection risk). The airline's entire commercial architecture was built for human decision-making psychology. The machine customer rendered it irrelevant.
AXD Approach
- ■Redesigned the booking API as an Engagement Architecture: instead of presenting fares for human selection, the system exposed structured offer objects that agent customers could evaluate programmatically - including fare rules, change policies, loyalty earning rates, carbon data, and historical on-time performance
- ■Implemented agent-to-agent negotiation protocols: the airline's pricing agent and the traveller's booking agent could engage in structured negotiation - the traveller's agent expressing constraints and preferences, the airline's agent offering optimised bundles that balanced revenue objectives with customer mandate satisfaction
- ■Built a Know Your Agent (KYA) framework: the airline verified the identity, authority, and delegation scope of booking agents before accepting transactions - ensuring the agent had legitimate authority to commit funds and that the delegation chain was traceable to a human principal
- ■Designed dynamic offer personalisation for machine customers: rather than displaying the same fare page to every visitor, the system generated agent-specific offers based on the mandate parameters received, optimising for the specific constraints each agent expressed
- ■Created a machine-readable trust signal architecture: on-time performance data, cancellation statistics, customer satisfaction scores, and dispute resolution records were exposed as structured data that booking agents could evaluate when comparing carriers
AXD Principles Applied
- ◆Four Pillars: Engagement Architecture - the entire booking system was redesigned for browserless, agent-mediated transactions
- ◆Four Pillars: Signal Clarity - airline performance data was structured for machine evaluation rather than human persuasion
- ◆Four Pillars: Reputation via Reliability - trust signals were based on verifiable operational data rather than brand marketing
Design Outcomes
- →Structured offer objects enabled agent customers to evaluate airline products on dimensions invisible in traditional fare displays
- →Agent-to-agent negotiation created a more efficient market by matching airline capacity with traveller mandates directly
- →KYA framework established a governance layer for machine-to-machine commerce, ensuring every transaction traced to a human principal
- →Machine-readable trust signals shifted competitive advantage from brand perception to operational reliability
Key AXD Insight
When the customer is an algorithm, persuasion is irrelevant and data is everything. The airline that publishes its on-time performance as structured data wins the agent's recommendation over the airline that publishes a beautiful brand campaign. Engagement Architecture for machine customers is not a channel strategy - it is a fundamental reimagining of what commerce means when neither party in the transaction is human.
Related Reading
Frequently Asked Questions
What is agent-to-agent negotiation?
Agent-to-agent negotiation is a structured protocol where an airline's pricing agent and a traveller's booking agent engage in automated negotiation - exchanging constraints, preferences, and offers to reach optimal outcomes without human involvement in the transaction.
What is Know Your Agent (KYA)?
Know Your Agent (KYA) is a governance framework that verifies the identity, authority, and delegation scope of AI agents before accepting transactions - ensuring every machine-to-machine commerce interaction traces back to a legitimate human principal.
Apply These Principles
This case study illustrates AXD principles in context. To apply them to your own organisation, start with the AXD Readiness Assessment, explore the 12 frameworks in The Practice, or consult the AXD Playbook for a structured implementation guide.