Agentic Commerce

Core Thesis

Agentic commerce = commerce where AI agents act on behalf of users to discover, evaluate, negotiate, and purchase goods/services. Agents replace or augment human decision-making in the procurement process.

Key Transformation: Information Matchmaker Erosion

Traditional commerce relies heavily on information matchmakers:

Search engines (Google)
Marketplaces (Amazon, Taobao)
Comparison sites
Product reviews aggregators

Agent era disrupts this: Agents collect info directly, compare at scale, and don't need "discovery" in the human sense.

Why This Matters

Attention freed: Agents have no human attention bottleneck
Collection cost near zero: Agents gather extensive supplier data
Evaluation at scale: Compare hundreds of options simultaneously
Trust shifts: verify-not-trust replaces brand trust

The Procurement Manager Analogy

Jimmy developed this framework during the research:

Human procurement manager:

Limited bandwidth (attention)
Relies on reputation, relationships, certifications
Makes decisions with incomplete information
Vulnerable to sales/marketing influence

Agent procurement manager:

No attention bottleneck
Verifies claims directly (no trust needed)
Collects comprehensive supplier data
Ignores advertising

Impact on Advertising

Traditional Ad Models (Designed for Humans)

CPM: Cost per 1000 impressions (human attention)
CPC: Cost per click (human action)
CPS/CPA: Cost per sale/action (human conversion)

Agent Era Problem

Agents don't see ads. Agents don't click. Agents don't have "impressions." Ad models designed for human attention don't transfer.

Emerging Patterns

Agent-readable ads: Structured data that agents can evaluate
Incentive-based: Providers offer better terms for agent-direct purchasing
Trust signals: Verifiable credentials replace brand marketing
Procurement-based: Similar to B2B procurement with negotiated contracts

Info Intermediary Evolution

See info-intermediary for detailed analysis.

Trust Mechanisms

See verify-not-trust for detailed analysis.

Real Examples (Near-term 1-year horizon)

Verified Not Trusted

Financial: Agent reads SEC filings directly, ignores analyst reports
Healthcare: Agent checks FDA approvals, ignores pharma marketing
Legal: Agent reads contracts, ignores legal marketing
Products: Agent verifies specs, ignores product reviews

Procurement Agent Winners

Tokens & Data APIs: Easiest to implement (digital goods, instant delivery)
Flights/Hotels: Direct API integration with loyalty programs
Financial services: Auto-comparison of rates, direct fulfillment