Google Says AI Search Is Still SEO. What That Leaves Out for Product Pages
- Author
- By Dor
- Publication date
- Published
- Last updated
- Updated
- Reading time
- 5 min read
You already have an SEO agency and a Semrush seat, and someone has asked whether you need a second program for AI search. Google's answer is no. Its AI search optimization guide, published on developers.google.com in May 2026, says that from Google Search's perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO. That is right for Google's index. It is incomplete for the product data that ChatGPT and Merchant Center read.
The guide uses two industry terms worth defining once. Answer engine optimization (AEO) and generative engine optimization (GEO) are labels vendors coined for showing up in AI generated answers. Google folds both into SEO. We use a plainer phrase, AI visibility, for the outcome a Head of Ecommerce actually cares about: whether your products appear when a shopper asks ChatGPT or Google AI Mode what to buy.
What Google's guide says
The guide is short and unusually direct. AI Overviews and AI Mode are rooted in Google's core ranking and quality systems, using retrieval-augmented generation and query fan-out against the existing index. The advice that follows is foundational SEO: pages that are indexed and snippet-eligible, non-commodity content with real expertise, semantic HTML, good page experience, and less duplication.
The Mythbusting section is the news. For Google Search, the guide says you can ignore:
- llms.txt files and other special markup. Google Search does not use machine-readable AI text files, markup or Markdown to decide what appears.
- Chunking content into small AI-friendly pieces. Google's systems can pick the relevant passage from a page covering several topics.
- Rewriting content for AI, such as chasing every long-tail keyword variation.
- Seeking inauthentic mentions across blogs, videos and forums.
- Special AI schema. Structured data is not required for generative AI features, though Google still recommends it for rich result eligibility.
Two lines in the same document matter more for ecommerce than the myths. Google says that where appropriate, generative AI responses can include product listings and product information, and that Merchant Center feeds can help your products be visible in both AI responses and other Search results. It also points to the Generative AI performance report in Search Console for measuring it. Search Engine Journal's summary of the guide is a good companion read.
Why Google is right about Google
The still SEO framing is accurate for Google's own surfaces. AI Overviews and AI Mode are a generation layer on top of the same crawl, index and rank stack. The content quality, authority, structured data and topical depth that earn organic visibility also earn citations in the AI answer, and Google AI Mode shopping results draw on the same Shopping Graph as the rest of Google.
So if your agency is doing solid work on category and product pages, most of it carries into Google's AI surfaces. Product schema markup still matters for product cards and rich results. Page quality still decides which sources Google's query fan-out selects. Nobody needs to buy an AI-specific rewrite of the pages that already rank.
Where the guide stops
The qualifier is in the phrase from Google Search's perspective. Google's guidance covers Google's index. It says nothing about how other assistants find and describe products, and it says little about the product information itself, which is the part most ecommerce SEO programs never touch.
- Google AI Overviews and AI Mode read Google's index and the Shopping Graph. Classic SEO plus a complete Merchant Center feed covers it.
- ChatGPT accepts product data from merchants directly. OpenAI's product feed specification documents the fields it takes, including price, availability, category, material, color, size and a Google-compatible upload path, and its shopping research looks across the internet for price, availability, reviews, specs and images.
- Merchant Center reads the feed, not the page. Google's AI performance insights report lists the conversational terms shoppers use and the structured attributes, such as size, color and material, that may be missing from your product data.
The common thread is product information. A product page that never states the width of a boot, a feed with a stale price, or a category Google guessed wrong will be left out of an AI answer on every surface, and no amount of page-level SEO fixes it. The structured product data your team maintains serves Google Search, Merchant Center and ChatGPT feeds at once, which makes it the cheapest place to work.
What this means if you already have an SEO agency
Keep the agency. The work it does on crawlability, content and links still counts for Google, including AI Overviews and AI Mode, and Google has now said so in writing. Our ecommerce SEO work sits alongside that kind of program rather than replacing it.
The gap is in two places your SEO retainer rarely covers. First, the product data in Merchant Center and in a ChatGPT feed: complete attributes, current prices, correct categories, and the question and answer content that lets an assistant answer a specific shopper. Second, the shopper questions themselves. Most teams we meet say they only check whether they show up in ChatGPT when someone happens to ask, and nobody owns the list of questions a real customer would type.
That second part is what AI visibility work adds: a set of realistic shopper questions for your category, a record of whether your products appear as a product card, a mention, or not at all in ChatGPT and Google AI Mode, and the product data fixes that move the answer. It is measured before and after, on your products, so you can see what changed.
Three decisions for a Head of Ecommerce
- Do not fund a separate AI SEO program for Google. Ask your agency to confirm the product and category pages are indexed, snippet-eligible and carrying product schema, and to report the Generative AI performance data from Search Console.
- Give the Merchant Center feed an owner outside paid media. Review the AI performance insights report for missing attributes and fix them at the source, so they stay fixed.
- Decide who owns the shopper questions. Pick 20 questions a real customer would ask in your category and check whether your products appear in ChatGPT and Google AI Mode. If you want a starting point on one product, the free AI readiness report reads a product page and shows what an AI answer can learn from it and what is missing.
Frequently asked questions
Is AI visibility only SEO now that Google said so?
For Google's own surfaces, largely yes. For ChatGPT and other assistants that take product data directly from merchants, the inputs are different, and the product information in your feed and on your page does most of the work. The sensible plan is one SEO program plus an owner for product data and shopper questions.
Should my team stop working on structured data?
No. Google's guide says structured data is not required for its AI features but still recommends it for rich result eligibility, which is how product cards with price and availability appear. The same facts in your schema should match your feed and your page.
How do I know if the work is paying off?
Track three numbers separately: appearances in AI answers for your list of shopper questions, Merchant Center status for the products you fixed, and sessions referred from AI assistants. We walked through the analytics setup in how to measure AI referral traffic in GA4 and Shopify.
Google's guide settles one question: you do not need a second SEO program for AI Overviews and AI Mode. It leaves open the one your SEO agency is not looking at, whether your product data answers the questions shoppers actually ask in ChatGPT and Google AI Mode. A Paz expert finds those gaps in your product pages and feeds, you approve the fixes, and the platform publishes them and tracks whether you show up. If that is the question on your desk, talk to us.
How AI-ready are your products?
Check how ChatGPT, Google AI, and Perplexity evaluate any product page. Free score in 30 seconds.