20% of ChatGPT Conversations Are Shopping. Most Brands Aren’t There.

Key Takeaways

• 20% of ChatGPT chats now contain shopping intent (~180 M weekly)
• Only ~20% of merchants have machine‑readable product feeds
• Product cards, driven by 30+ structured attributes, boost conversion versus brand mentions
• Multi‑platform AI visibility requires feeds for ChatGPT, Google AI Mode, Perplexity
• Tools like Paz.ai can measure found rate and competitive ranking across AI agents

Pulse Analysis

The emergence of generative‑AI assistants as shopping intermediaries is no longer speculative. OpenAI’s latest data shows that one in five ChatGPT interactions now carries purchase intent, translating to roughly 180 million weekly shopping‑related conversations on a platform with 900 million active users. This volume dwarfs many traditional e‑commerce traffic sources and mirrors the scale of early mobile commerce booms. However, the AI‑driven funnel differs fundamentally from keyword search: shoppers describe needs in natural language, and the agent returns product cards only when it can match a richly described SKU.

The decisive factor is the quality of the product catalog. AI agents ingest machine‑readable feeds—XML, JSON‑LD, or Shopify’s native export—and rely on explicit fields such as material, dimensions, color variants, price, shipping, return policy, and contextual use‑case tags. Brands that expose 30 or more attributes per SKU, like Spanx and Glossier, consistently appear as clickable cards, while those with sparse metadata are relegated to mere brand mentions. Traditional SEO signals like title tags or meta descriptions have little influence on these agents; structured attributes are the primary matching signals.

For retailers, the path to capture AI‑shopping demand is straightforward yet urgent. First, audit the attribute depth of every SKU and enrich feeds to meet the 30‑attribute benchmark. Second, deploy visibility monitoring tools—Paz.ai, Feedvisor, or similar—to track found rates, card appearances, and rank positions across ChatGPT, Google AI Mode, and Perplexity. Finally, iterate by testing high‑intent queries and refining contextual fields, turning AI mentions into revenue‑generating product cards. Brands that act now can tap into a channel already delivering hundreds of millions of purchase intents each week.