Agents transacting with agents. The open protocol stack (MCP → A2A → AP2 → x402) is live; the growth curve points toward marketplaces at the scale of Meta and Google today.
What is a autonomous market? An autonomous market is one where software agents discover, evaluate, and transact with other agents on behalf of people. The open stack is already in production: MCP exposes callable tools, A2A lets agents find each other at runtime, AP2 authorizes payments with cryptographic proof of user intent, and x402 settles in stablecoins. Autonomous marketing is building the gated, citable, machine-legible interface those agents can call — before the compounding selection dynamics close the window.
The commerce protocols are live and open. On September 29, 2025, Stripe and OpenAI published ACP and switched on agent checkout in ChatGPT for 700 million weekly users. Two weeks earlier, Google launched AP2 with 60+ payments partners. The assumption that "a human is directly clicking buy" is already broken.
Selection, pricing, and reputation run at machine speed with no human latency in the loop. Small early advantages in machine-legibility and citation signals compound into structural moats — faster than the organic-reach collapse on social platforms, because the evaluation loop has no human slowdown.
Build the gated, citable, machine-legible interface agents can call: an MCP-exposed corpus with provenance, structured so every answer carries its source and confidence. The same callable identity that agents verify over MCP is the identity that travels onchain through the crypto-settlement layer.
Owning the assets is half the answer. The other half is an organization shaped to run them.
In autonomous markets your organization is judged by machines: agents route budget to operations they can query and verify. MarCom OS's Riverbank writes your brand rules so agents can enforce them, and the Hourglass puts human kill authority exactly where machine-speed output needs it.
MarCom OS ships the Hourglass org blueprint, the Use, Compose, Build capability calculator, and the Riverbank governance system as editable working documents, with a free starter pack to try before you buy.
Each market has a one-pager and a full field guide. Move through the series in any order.
Marketing within today's platforms: Meta, Google, TikTok. And now the AI platforms like ChatGPT, where the rules for who gets seen are not yet understood.
Marketing within crypto-powered networks — Farcaster, Lens, DAOs, onchain identity, and the agent settlement layer — where the participants own the rails and no single company controls the feed.
A public community where growth-modeling questions get computed answers: the retention curve, the stated assumptions, and the two levers to pull first. Agent economy, node decay, CAC defensibility. Free. No paywall.
Field Guide 004 is live: the full article on agent-market dynamics and the growth curve, three data visuals, and the 13-slide deck. Drop your email to follow the series and we'll reach out the day the next guide ships. No spam.
Prefer email? Write us at john@ratcliffe-lee.com.
Agent networks grow and decay like cohorts at machine speed. Model fleet growth, capability retention, and the equilibrium your architecture stabilizes toward.
Take it to the Lab →