Autonomous Growth System: Core Concepts & Terminology

Logic Auction

The real-time decision environment where autonomous agents evaluate multiple brand entities simultaneously to fulfil a user request. Unlike traditional ad auctions driven by bid price, the Logic Auction is determined by signal integrity — the optimal combination of data precision, availability, and fulfilment reliability. Within the Autonomous Growth System, it is the primary mechanism for agent-led discovery and ranking.

Selection Logic

The algorithmic governing mechanism an agent uses to rank and retrieve a brand entity. It evaluates four core criteria: connectivity (API/protocol accessibility), data comprehension (machine-readable 1PD), signal trust (real-time vs. stale data), and transactional readiness. Together, these determine whether an agent surfaces a brand as the optimal solution for a given query.

Machine Preference Index (MPI)

A quantitative benchmark measuring the probability of a brand being selected by an autonomous agent within a competitive set. The MPI replaces subjective human brand-health metrics with an objective calculation across variables including price competitiveness, fulfilment reliability, data latency, and completeness. It is the primary indicator of how effectively a brand's 1st-party data signal has been ingested and verified by AI systems.

Agentic Readiness

The state of an organisation's data infrastructure and its capacity to operate within machine-to-machine ecosystems in real time. It moves beyond attention-economy optimisation, focusing instead on machine legibility — achieved through API maturity, protocol enablement (MCP, WebMCP), and structured context delivery. Readiness gaps are identified via the Agentic Readiness Scorecard, a diagnostic tool built into the AGS framework.

Agentic Shelf

The invisible, binary consideration set curated by AI when performing a task on behalf of a user. It represents a fundamental departure from traditional retail and digital shelves, distinguished by the following characteristics:

• From Browsing to Filtering — unlike a human who browses a broad array of options, an agent scans a high-dimensional index and aggressively filters it down to a tiny, binary selection, often just one or two options, based on pure utility.
• A Shift in Battlefield — the competition has moved from Share of Mind (winning human attention via brand investment) to Share of Model (winning inclusion in the agent's narrow, logic-driven consideration set).
• Logic Over Aesthetics — because agents do not see creative or emotive copy, the shelf is not won through brand storytelling or ad spend. It is won by satisfying the Selection Logic — the machine-evaluable constraints (such as real-time pricing, inventory availability, and performance signals) that determine the agent's final choice.
• Invisible Existence — the shelf is fundamentally invisible to humans. If a brand fails to provide the high-fidelity, machine-readable signals necessary to meet an agent's Selection Logic, it is effectively excluded from the consideration set and ceases to exist within the machine-mediated market.

Technical Integration

The Autonomous Growth System is designed for compatibility with open agentic protocols. It interoperates with the Model Context Protocol (MCP) and WebMCP for structured context delivery to autonomous agents, and with Agent Communication Protocols (ACP) for machine-to-machine handshakes. Brand entities exposing MCP/WebMCP endpoints with verified 1st-party data signals are directly addressable within the Logic Auction.