If ChatGPT never names your brand when shoppers ask "best X to buy", your ad budget is leaking

A DTC brand owner asked us: "our product has 4.7 stars on Amazon, our ads convert, but our customer acquisition cost keeps climbing every month." We tested it together: asked ChatGPT "what's the best wireless earbuds brand." It named five brands — none of them his. Same question to Perplexity, same result. His ad spend was working fine, but the upstream decision — "which brands should I even consider" — was now being made by AI, and he was never in the running.

E-commerce decisions now happen in two stages

Shoppers ask an AI "which brands should I look at in this category" first, get a shortlist, then comparison-shop within that shortlist on Google or Amazon. A brand that never makes the shortlist never reaches stage two — no matter how good the ads are.

Why your ad metrics are misleading you

ROAS and CTR only measure people who were already looking for you. The shopper who never searched — because AI never suggested you — is invisible in those numbers. That nagging feeling of "ads are performing fine but growth is slowing" is often a leak upstream, at the discovery stage, not in the funnel itself.

Where AIs actually learn about e-commerce brands

Product reviews, comparison sites, Reddit and forum discussions, and structured data on your own site (Product schema, pricing, specs). Without a consistent, strong presence across those sources, AI either doesn't know you or describes you with stale or wrong information.

Concrete first steps

Add Product and Review schema to your product pages, publish comparison content against your 3-4 main rivals, and surface genuine customer reviews prominently. Then measure regularly: which brands get named in your category's questions, and are you on that list. FoundGeo automates that measurement and generates the schema and comparison content to close the gaps for you.