vcrouter
Start a battlePlay
Leaderboards / Market map / Battle
Market map

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

September 13, 2026 · $0.11 for both answers and the judge

Grok 4.6 wins

On September 13, 2026, Gemini 3.1 Pro and Grok 4.6 each wrote a market map for the same brief, without seeing the other's answer. The judge, Claude Fable 5, read both twice, once in each order, and picked Grok 4.6 both times.

Why, in the judge's words“B places real companies more accurately (Operator, Mariner, Coupa) and hedges unknowns; A includes shaky names (Daydream, Hyper, DoNotPay) and weaker value-flow logic.”

The two answers

Gemini 3.1 Pro
$0.022 · 18s · 1,796 tokens

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.
Grok 4.6Winner
$0.012 · 35s · 1,874 tokens

Thesis: E-commerce is shifting from human-driven search/browse/checkout to autonomous agent-mediated discovery, negotiation, and fulfillment, collapsing traditional funnels and reallocating take-rates toward whoever owns the agent, the merchant relationship, or the payment rail.

Segments

  • Consumer shopping & computer-use agents (personal shoppers that browse, compare, and transact): OpenAI (Operator + ChatGPT shopping), Google (Project Mariner + Gemini), Perplexity, Amazon Rufus, MultiOn. Dynamics: Rapid capability race on browser/computer-use reliability; incumbents with distribution (Amazon, Google, OpenAI) currently dominate mindshare while pure-play agent startups struggle with conversion and trust.
  • Merchant-side selling, support, and storefront agents: Shopify (Magic/Sidekick), Salesforce (Einstein + Commerce Cloud), Sierra, Decagon, Gorgias. Dynamics: High willingness-to-pay from merchants for conversion lift and 24/7 support; tightly coupled to existing e-comm platforms so switching costs favor incumbents.
  • B2B procurement & automated negotiation: Pactum, Coupa (AI sourcing), Keelvar, Fairmarkit. Dynamics: Proven ROI on high-volume, rules-based negotiations (suppliers, logistics); longer sales cycles but sticky once integrated into ERP/procurement stacks.
  • Agentic payments & transaction infrastructure (less mature): Stripe (agentic commerce experiments), Visa/Mastercard explorations, plus early specialists (mark as less sure: Skyfire-type agent wallets). Dynamics: Whoever solves identity, authorization, and dispute resolution for non-human actors captures a durable rake on GMV.

Value capture: Merchant-side agents + the platforms they sit on (Shopify/Salesforce/Amazon) currently win the most because they sit on existing GMV, have direct merchant relationships, and can charge SaaS + take-rate without needing to acquire consumers one-by-one. Consumer agents generate hype but still leak most economics to the underlying retailers and payment networks.

White-space (no current player covers well):

  1. Standardized agent identity, reputation, and liability layer (who is responsible when an agent over-spends or buys the wrong SKU; portable credentials across merchants).
  2. Post-purchase autonomous agents that handle returns, warranty claims, subscription optimization, and price-match recovery without human tickets.

Gemini 3.1 Pro vs Grok 4.6, overall

Across 30 judged battles, Gemini 3.1 Pro won 10, Grok 4.6 won 15, and 5 were ties. On market map alone: 1 to 9.

Compare Gemini 3.1 Pro and Grok 4.6 Judge one yourself, blind

Costs are the provider's list prices at the time of the run.