Product data: the part of AI visibility you control
Product data is the only part of AI visibility a merchant fully controls. These pages cover the rules AI shopping surfaces publish for titles and attributes, how to write a title an assistant can match to a real question, and what meta descriptions and FAQ schema still do now that answers are generated.
- Start here
The product data rules AI shopping surfaces actually publish
OpenAI's product feed spec sets hard limits, including a 150-character title maximum. What the published rules say, and how to check your catalog against them.
The OpenAI product feed spec, explained for Shopify merchants
The published spec behind ChatGPT shopping: required fields, the is_eligible flags, the 150-character title rule, and which Shopify admin field each one maps to.
How to write product titles an AI assistant can actually use
A product title tuned for Google is too short for AI shopping. What to include, how long to go, and the before-and-after pattern that fixes most catalogs.
Why a product called Summer Dress is invisible to AI
Shoppers ask AI constrained questions: material, size, budget, use case. A product without attributes cannot match a constraint. The completeness checklist, from the published specs.
Product descriptions AI can quote: substance over sales copy
Assistants quote facts, not adjectives. How to write product descriptions inside the 5,000-character spec limit that answer real buyer questions, and why duplicate supplier copy fails twice.
Do meta descriptions still matter now that AI answers the question?
Meta descriptions were never a ranking factor and are not an AI shopping field. They still control your snippet and still earn clicks. What they are for.
Is FAQ schema still worth adding after Google retired FAQ rich results?
Google fully retired FAQ rich results in May 2026. FAQPage markup no longer produces rich snippets. Why structured answers still matter for AI retrieval.