[Two Cents #80] “Flights of Thought” on Consumer + AI — Part 6: Commerce — 2. GEO for Shopping, Headless Commerce
Introduction
It’s becoming clear that the market’s “readiness” for Consumer AI has crossed a tipping point.
What we need now is to get far more concrete about how AI-driven market change will unfold—the direction, the mechanisms, and the implications for industry structure, competitive dynamics, and the economics between participants.
For founders, the job is to identify those opportunities a little earlier and move first. For investors, the job is to recognize those early moves quickly and support them aggressively.
This series—my “Flights of Thought”—is an attempt to share how I’m thinking through what will happen, what it will unlock, and what kinds of ideas are likely to matter.
Now it’s time to move from macro themes to concrete opportunity spaces by vertical.
I’ll start with commerce, because it’s arguably the clearest category to model—at least on paper. (That confidence may prove wrong, but it’s a good place to begin.)
The beginning of the end
For the last two decades, online commerce has followed a surprisingly stable customer journey. A consumer searches Google a handful of times, visits a dozen sites to compare information, and eventually checks out on one specific store. (Fashion is the exception—more browsing-oriented, closer to scrolling Instagram than “research then buy.”)
In that world, the monetization stack is clear: Google captures search ads, Amazon captures sponsored listings, and creators capture affiliate economics.
AI is now breaking that loop.
First, the purchase journey is starting to close inside AI chat surfaces—ChatGPT, Perplexity, and others—where discovery, recommendations, browsing/comparison, and even checkout can happen end-to-end in a single interface.
Second, if agentic commerce (discussed in [Two Cents #79]) becomes more common, the legacy funnel that begins with Google search will compress further—often into agent-to-agent interactions that don’t require a “search step” at all. The mechanics of that shift tie back to the agent framework in [Two Cents #77].
In the near term, the market impact is showing up through two distinct pathways:
the rise of GEO, and
the rise of headless commerce.
The rise of GEO
The first phase keeps the current shopping platform structure largely intact—but changes how intent-driven traffic reaches those platforms.
The practical shift is simple: the consumer-brand interface moves from web pages to AI conversations. Marketing strategy has to move with it—from “drive traffic to our site” to “make sure ChatGPT recommends us first.”
That optimization game is GEO (Generative Engine Optimization). It’s the evolution of SEO from keyword/backlink mechanics tuned for Google’s ranking algorithm to structures and signals that AI chatbots and AI agents can understand, trust, and recommend.
You can already see the early motion:
Google itself is reshaping discoverability via AI experiences.
A first wave of GEO startups (e.g., Profound and others) is emerging to help brands adapt.
The next phase will look different. As agentic commerce becomes more common, “GEO for chatbots” won’t be enough. We should expect a more agent-native version of GEO—optimized for how agents request data, compare options, and transact.
These two GEO arcs—chatbot-focused vs. agent-focused—will evolve differently. Which one dominates depends on where most consumer purchase journeys end up:
closed-loop “chat commerce,” or
delegated “agent commerce.”
My working expectation: chatbot GEO leads in the short term, because it’s already here; agent GEO dominates in the long term, because I’m skeptical that a CLI-like chatbot interface remains the primary consumer UX forever (as discussed in [Two Cents #76]).
The rise of headless commerce
In parallel, there’s a second structural change: if chatbots/agents route consumers directly to product pages (or bypass front-end browsing entirely), the traditional ecommerce front-end becomes less important. Commerce becomes headless.
We already saw the shape of this in early GPT-4 demos—function calling that completes an Instacart order or books travel via Kayak. The interface is no longer “the store.” The interface is the AI. The store becomes an API.
In headless commerce:
storefront UX matters less,
product data and checkout APIs matter more.
This is why incumbents are taking different stances on how much access external agents get—product data, inventory, pricing, checkout. As argued in [Two Cents #79], platforms with “default destination” power (Amazon-like) see headless access primarily as a threat, while platforms that compete for distribution (Shopify-like) can see it as opportunity.
What changes, concretely?
The GEO pipeline will likely split into multiple lines:
We move from optimizing metadata/keywords/backlinks for Google to creating structures that:
make it easy for LLMs to include products in generated answers (and surface them during AI-assisted web search), and
support discoverability and retrieval patterns for agents acting on behalf of intent-driven buyers (agentic-commerce GEO).
The affiliate/creator layer—historically an intermediary between search and product discovery—will need a major reset in role and economics.
New transaction networks will emerge to enable commerce discovery and placement inside AI interfaces—effectively ad networks and deal rails built for chat/agent surfaces.
In a headless flow, product data (price, inventory, options) and checkout are exposed via API or A2A interaction, while the consumer completes the journey inside an AI interface. The center of gravity moves away from “shopping UX” and toward “data + permissions + execution.”
Where the market structure can go
If you extend this logic, a few futures become plausible:
1) A “super-default destination” outcome
A small number of default destinations (Amazon-class) evolve into super-platforms by using agents to reach other platforms’ inventories—consolidating distribution even further. Call this the “unification” path.
2) A “democratized product access” outcome
Access to product information across Amazon/Shopify/long-tail merchants becomes more open and standardized. In that world, a new kind of default destination could emerge—one built specifically for agentic commerce.
Both paths create demand for new middle layers:
Product data aggregation layers that normalize access across fragmented platforms—Plaid-like infrastructure for commerce catalogs and availability.
Checkout layers that support delegated purchase flows: agent-initiated payments, user-confirmation payments, and agent-to-agent microtransactions—Stripe-like infrastructure for agentic commerce.
Tooling layers that connect, optimize, and measure the new system: data sync, ranking optimization, analytics, attribution, and policy enforcement.
Startup opportunities
This is still early, and the opportunity set will expand. But even now, a few categories are visible.
1) A2A ad/marketing networks + a new affiliate/attribution layer
The old model was simple: pay Google to buy position. In an agent world, new transaction patterns become possible:
Buyer-paid research: consumers (or their agents) pay a fee for high-quality information and comparisons—especially for high-consideration categories (lifestyle, cars, homes, “life purchases”). This can generalize.
Seller-to-buyer-agent incentives: seller/brand/affiliate agents bid—potentially via reverse auction—by explicitly offering incentives to the buyer’s agent to be preferred.
Generalize this and you get:
“Google Ads for agents”: preferred placement in agent decision loops, and
new attribution protocols: tracking “who influenced the agent’s purchase decision,” not just who got the last click.
2) Agent platforms for merchants — “Shopify 2.0 for agents/headless”
Merchants may run their own brand agents directly, rather than relying on platform storefronts.
Opportunities include:
Shopify accelerating into this and winning by default, or
a new “Shopify for agents/headless” competitor emerging, and
marketplaces that aggregate headless merchants and expose them to AI agents as the primary discovery channel.
3) Creator economy 2.0
Creators historically sat between search and product pages via affiliate networks. That intermediary role will be redefined.
Creators might:
be partially replaced by agents,
become agents themselves (brand-agent-like), or
operate “creator agents” that carry their voice, taste, and credibility into AI surfaces.
Possible platforms:
“Linktree for agents” (creators operating their own agent endpoints), and
discovery platforms for creator/agent entities (another “Shopify for agents/creators” layer).
4) Agent-optimized product information infrastructure + a new commerce aggregator/platform
We’re already watching platforms diverge on agent access policies. If product information becomes meaningfully unbundled from incumbent storefronts, the path from consumer to product changes fundamentally.
A useful analogy: the web unbundled access to information and democratized distribution. A similar unbundling could happen for commerce product information.
If that dynamic accelerates, the prize is enormous:
a new “head” (new default destination) built for agent commerce that aggregates product databases across platforms, and
the enabling infrastructure: “Plaid for Agentic Commerce.”
This is the category where the long-term map of commerce could look different a decade from now.
Market size and why this matters
Google is still the front door for a large fraction of online shopping journeys. In 2024, Google’s advertising revenue was ~$274B, and shopping-intent categories (e-commerce, retail, travel, local) are often estimated at ~40–60% of that. That implies $100B–$170B of annual ad spend tied directly to product discovery and comparison.
Amazon also dominates a massive “shopping search ads” pool. In 2024 it recorded roughly $56B in ad revenue, largely from sponsored listings closely linked to purchase intent.
Beyond traditional SEO and ads, affiliate/creator commerce has become a meaningful market:
global affiliate marketing spend in 2023: roughly $15B, and
influencer-driven affiliate commerce: roughly $20B (estimated).
Add those together and you’re looking at $180B–$250B per year of commerce-related SEO/ads/affiliate/creator economics that could migrate toward AI-native surfaces and GEO-native rails.
That alone is a “next Google Ads / next Amazon Ads” scale opportunity.
More speculatively (but directionally important): if these shifts change the balance of power among commerce platforms themselves—not just SEO → GEO—then the restructuring could reach much deeper into industry structure across an ecommerce market that’s ~$6T today, with long-term TAM discussions often in the $20T–$30T range. Full restructuring would likely take 10–20 years, but the early phases can move fast.
Closing
SEO defined the fate of online businesses for a generation. Ranking meant revenue. Entire ecosystems of tools and agencies grew around it.
Over the next decade, that center of gravity shifts toward AI-native GEO. In parallel, GEO + agentic commerce can change not only marketing economics, but the competitive dynamics between commerce platforms themselves.
And whenever an industry’s distribution and monetization rails get rewritten, startups get an unusually large window to build category-defining companies.
Call for Startups
The purpose of sharing this thinking is straightforward. As an early-stage investor focused on Consumer + AI, I hope this series helps existing startups better leverage AI-driven shifts—and helps new founders reduce trial-and-error as they search for meaningful opportunities.
In that sense, this is Two Cents’ version of a Call for Startups.
If you are an early-stage founder or startup in Consumer + AI and believe you are onto something, my inbox is always open. Feel free to reach out via DM or email:
hur at hanriverpartners dot com