AI-native commerce: agents that shop, sell, and negotiate

Here is an end-to-end venture market map for AI-Native Commerce.

1. The Thesis

AI is restructuring commerce from a "search-and-click" paradigm to an "intent-and-delegate" model, where autonomous agents intermediate transactions, dynamically negotiate pricing, and execute payments machine-to-machine (M2M) with minimal human intervention.

2. Market Segments & Dynamics

Segment 1: Autonomous Buyer Agents (The Shoppers)

• Companies: MultiOn, Daydream, Hyper (Hyper.space), DoNotPay, Amazon (Rufus).
• Dynamics: This segment faces high consumer friction regarding trust and autonomous spending limits, but holds massive top-of-funnel potential as agents evolve from mere recommendation engines to autonomous executors.

Segment 2: AI Sales & Support Agents (The Sellers)

• Companies: Sierra, 11x.ai, Decagon, Intercom (Fin), Shopify (Sidekick).
• Dynamics: Experiencing the fastest enterprise adoption because the ROI is immediately quantifiable through customer service headcount reduction and direct conversion rate optimization.

Segment 3: Algorithmic Procurement & Negotiation (The Dealmakers)

• Companies: Pactum, Vendr, Globality, Arkestro, Zip.
• Dynamics: A high-value B2B segment where agents autonomously negotiate vendor contracts and supply chain pricing; growth is strong but heavily gated by the ability to integrate with legacy ERPs (SAP, Oracle).

Segment 4: Agentic Financial Infrastructure (The Rails)

• Companies: Skyfire, Payman, Stripe (Agent Toolkit), Coinbase (CDP AI Wallets), Biconomy.
• Dynamics: The critical "picks and shovels" layer competing to establish the standard protocols for machine identity, agent wallets, and M2M value transfer.

3. Value Capture: Who Wins the Most Value and Why?

Agentic Financial Infrastructure (Segment 4) will capture the most value. Why: In an ecosystem where buyer agents and seller agents are constantly churning, competing on marginal UI/UX improvements, or becoming commoditized by foundational models, the infrastructure layer acts as the tollbooth. Whoever establishes the standard protocol for agent identity, spending limits, and M2M payment routing will take a fractional cut of every autonomous transaction, benefiting from massive network effects and avoiding the brutal customer acquisition costs (CAC) of the application layer.

4. White-Space Opportunities

• Agent-to-Agent (A2A) Marketing & SEO: Currently, brands optimize for human eyeballs (Google Ads, SEO, social). As buyer agents take over purchasing, there is no infrastructure for how a seller agent "pitches" a buyer agent. A white space exists for an A2A ad network or "Agent Search Engine Optimization" (ASEO) platform that formats product data specifically to influence the LLMs powering buyer agents.
• Agentic Liability Insurance & Automated Dispute Resolution: When an AI buyer agent gets tricked by a malicious AI seller agent (e.g., hallucinated specs, dynamic pricing manipulation), who is liable? There is a massive gap for automated arbitration protocols, smart-contract escrow services, or specialized insurance products designed specifically to underwrite and resolve M2M commerce disputes.