The Age of Agentic Commerce: How AI is Prescribing Products
This article is based on the presentation "Exclusive AEO Research from Profound's Zero Click NYC Event - October 2025," which revealed how AI is fundamentally restructuring the e-commerce funnel, shifting from merely recommending products to actively prescribing them and completing purchases.
The Invisible Commerce Funnel
The research, which analyzed 10,000 non-shopping-related ChatGPT conversations, uncovered that AI is now a commercial membrane between human problems and market solutions.
Unprompted Recommendations: Over one-third of conversations contained completely unprompted product recommendations [01:36].
Brand Prescription: In over 41% of those instances, the AI mentioned specific brands by name without the user asking for a product or a brand [01:52].
Problem to Product: The study concluded that people came to the AI with human problems (e.g., "My back hurts") and were immediately steered toward specific products, turning the AI into a powerful tool for instantaneous commercial verdict [02:08].
The Sarah Example: The presentation illustrated this with "Sarah," who went from complaining about back pain to purchasing an $1,800 Herman Miller Aeron chair based on a three-second AI recommendation she never researched herself. Her funnel was compressed from pain to purchase in just four hours [11:01].
The New Rules of AI Visibility
AI large language models (LLMs) are not reading the same internet as Google. Their decision-making process for content trust is rapid, trackable, and, crucially, influencable [02:46].
The 3-Second Process:
When a user submits a query (e.g., "A CRM for 50 employees"), the LLM goes through a rapid three-second process [06:28]:
Query Fan Out: The initial question is exploded into five to seven simultaneous searches (synthetic queries) to cast a wide net (e.g., "best CRM 2025," "small business CRM comparison") [04:16].
SER API Calls: The model instantly runs multiple real-time searches via a SER API, collecting 30 to 50 URLs in 500 milliseconds [04:37].
The Trust Decision: The model instantly disqualifies content based on AI-specific criteria and prioritizes other signals [04:48].
Signals that Disqualify Content:
JavaScript Heavy Sites: AI often skips pages that require JavaScript to load [04:57].
Affiliate Pages and Content Forms [05:05].
Slow Loading and Cryptic URLs [05:14].
Signals that Build AI Trust (The Five Pillars of AEO):
The key to AI Engine Optimization (AEO) is mastering the following signals, which can influence results in days, not months [06:59]:
Semantic URLs: URLs that clearly describe the content (e.g.,
best-CRM-for-startups) receive 11.4% more citations than generic links [19:23].Answer-Forward Content: The exact answer must be placed in a self-contained chunk at the top of the content. AI wants answers, not stories [19:33].
Genuine Community Discussion: Reddit is the number one most cited source for Perplexity and number two for ChatGPT. Real people talking beats marketing copy [05:40].
Freshness Beats Authority: Content published in the last 30 days receives exponentially more weight than older, even authoritative, content [05:49].
Plain HTML: Content that requires JavaScript to load (e.g., from a React app) is often invisible to the AI [19:54].
The Agentic Commerce Protocol
The speed of change in this space is unprecedented. In just 10 months (November 2024 to September 2025), the industry moved from AI recommending products to AI completing the purchases themselves [18:49].
Agentic Commerce Protocol: OpenAI and Stripe launched a protocol that allows ChatGPT users to buy directly from over a million Etsy and Shopify merchants [16:11].
Programming the Buyer: The shift means merchants are no longer just listing products; they are programming the AI's shopping brain [15:26]. Every product is becoming machine-readable, and every transaction is becoming programmable.
Exponential Success: Success in AEO creates a virtuous circle: a product recommended by AI on day one leads to more sales, which leads to more positive reviews. AI then weights those new, fresh reviews, recommending the product more strongly. This leads to that product owning the category over time [11:55].
The Stakes: AI already influences $104 billion annually in commerce, a number projected to grow exponentially [12:29]. The top 2,000 brands in the study capture nearly half of this value, leaving the remaining 200,000 brands to fight over a smaller, fragmented piece [12:43].
The presentation concluded with a decisive choice for companies: continue marketing to humans while AI mediates every decision, or start marketing to the machines that are now making the decisions [20:26].