Does FAQ content help ChatGPT recommend you?

Short answer: Not much. ChatGPT picks Shopify products from the Shopify Catalog record, which carries your description, options, price, images and policies, not your FAQ block. Store questions are answered from a separate endpoint fed by the Knowledge Base app. Octolift, built by Craftshift, scores the record that actually gets read. Free for 5 products.

“Add an FAQ section to every product page for AI” is the most repeated piece of AEO advice on the internet, and for Shopify product recommendations it is mostly wrong. Not harmful. Just aimed at the wrong reader. The three parties that actually decide whether your product gets recommended (Shopify, OpenAI and Google) have each written down what they read, and a product page FAQ block barely appears in any of it.

This is a short post because the sources are short. We’ll quote them, draw the line between the two places FAQ content can live, and tell you where the hour you were going to spend on FAQ markup would actually move the needle.

In this post

• Does FAQ content help ChatGPT recommend a Shopify product?
• Two paths, and FAQ content only touches one
• What Google says about FAQ markup and AI features
• Where store FAQs actually live
• When product FAQs are still worth writing
• What to do with the hour instead
• Frequently asked questions

Does FAQ content help ChatGPT recommend a Shopify product?

Start with what OpenAI says ChatGPT reads when it chooses products: “structured metadata from first-party and third-party providers (e.g., price, product description) and other third-party content”, the model’s own prior knowledge, and OpenAI’s policies. And for Shopify stores specifically: “product data is already integrated into ChatGPT through Shopify Catalog… No additional work is required from individual merchants.”

Now what Shopify Catalog carries. Per Shopify’s help centre, products are listed “with their title, description, options, images, price, availability” and other key attributes, and the catalog infers attributes, tech specs, top features and selling points from that. There is no FAQ field in the record. A question and answer block on your product page is not in the data ChatGPT’s shopping results are built from. It can only reach the assistant through the open web crawl, which is the second, noisier path.

So: does FAQ content help? Through the catalog path, no. Through the crawl path, a little, in the same way any clearly written text on the page helps a crawler. Which is a long way from “every product needs an FAQ block for AI”.

Two paths, and FAQ content only touches one

	Catalog path	Open web path
What the assistant reads	The Shopify Catalog record: title, description, options, images, price, availability, policies, plus inferred specs	Your rendered product page, like a search crawler
Who uses it for products	ChatGPT shopping results, Gemini, Copilot, Meta, via Agentic Storefronts	Any assistant doing a live web search
Does a product page FAQ block count?	No. It is not a catalog field.	As ordinary page text, yes; FAQPage markup adds little on top
What moves it	The five listing quality signals: description, images, reviews, variants, policies	Crawlable pages, clear text, structured data that matches the visible text

Why does this distinction get lost? Because most AEO advice is written for every website on earth, where the crawl path is the only path. Shopify stores have a second, structured path that the assistants prefer for shopping, and advice that ignores it sends merchants to polish the wrong surface. We explain the full mechanism in how ChatGPT chooses Shopify products to recommend.

What Google says about FAQ markup and AI features

Google is unusually blunt on its AI features page: “There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary.” And: “You don’t need to create new machine readable files, AI text files, or markup to appear in these features.” And, in case that left room: “There’s also no special schema.org structured data that you need to add.” What it recommends instead is the ordinary list: allow crawling, link to your pages internally, and make “sure your structured data matches the visible text on the page”.

FAQPage markup also lost its old day job. Google removed FAQ rich results from Search in May 2026, so the markup no longer earns the expandable snippet it used to. It still describes your page’s structure to anything parsing it, and if you have a genuine FAQ block the markup should match it (that’s Google’s “matches the visible text” rule). But adding FAQ schema to get into AI answers is adding a label to a box the assistants don’t open.

(A note for the people who ask: our free JSON-LD schema generator still produces FAQPage markup, because a visible FAQ with matching markup is correct. It just isn’t a visibility lever, and we’d rather tell you that than sell you a feature.)

Where store FAQs actually live

Here is the part the generic advice misses entirely. When a shopper asks an assistant “do they ship to Canada” or “what’s the return window”, the assistant does not read your product page or your FAQ page. Every Shopify store exposes a public endpoint with a tool Shopify documents as search_shop_policies_and_faqs, which “answers questions about the store’s policies, products, and services” with “no authentication required”. It answers from your admin settings and policies, and from the FAQs stored by Shopify’s free Knowledge Base app, and it returns the sources it used.

That is where store FAQ content belongs: in the Knowledge Base app, drawn from your refund policy, shipping policy and contact page, so the assistant’s answer about your store matches what your pages promise. It is also the one place where “writing FAQs for AI” is literally true. The whole method, including the ten questions assistants ask most, is in Shopify Knowledge Base: the FAQs AI agents read.

Notice what that endpoint is for, though. Accuracy, after the product has been found. Shopify’s own framing of the Knowledge Base app is that it customises the FAQs agents use to answer questions about your store; it is not a discovery lever. Which brings us back to the question in the title, and to the honest answer: FAQ content makes the assistant right about you. It does not make the assistant pick you.

When product FAQs are still worth writing

• When the answers are facts you’d otherwise leave out. “Does it fit a 15-inch laptop?” is an FAQ. “Fits laptops up to 15 inches in the padded compartment” is a description sentence, and the description is in the catalog record. Write it in the description first; the FAQ block can repeat it.
• When humans scan. Shoppers read FAQs. A good one lowers support load and returns. That is a conversion reason, not an AI reason, and it’s a fine reason.
• When you sell to non-Shopify channels through the crawl path. An assistant doing a live search of your page reads whatever is on it. Clear question and answer text helps there, markup or not.

And one rule in all three cases, learned the hard way when we built the FAQ drafting side of Octolift: every answer has to be supported by text already in the listing or your policies. We made the drafter quote the exact passage it drew from and refuse anything it couldn’t cite. An FAQ that invents a fact (“yes, it’s waterproof”) is worse than no FAQ, because an assistant will repeat it with confidence.

What to do with the hour instead

Open Sales channels, Agentic in your admin and look at the five listing quality signals on your best sellers: description completeness, image coverage, product reviews, variant and option completeness, shop policy completeness. Those are the fields the catalog record is built from, and Shopify grades them per product. Then do the store side once: ten Knowledge Base answers from your policies. That hour beats any amount of FAQ markup, and Shopify’s five listing quality signals for AI walks through each signal by hand.

If you’d rather see all of it scored at once, Octolift reads the live Shopify Catalog record for every product, scores it out of 100 and drafts the description fixes, and its Knowledge base page measures the store endpoint against the ten questions. Free for 5 products, $29 a month for 100. The scoring guide lists the twelve checks, and FAQ schema is deliberately not one of them.

Two more things FAQ advice tends to skip. Blocking AI crawlers in robots.txt affects only the open web path, and our rubikify.com post on whether to block AI bots is the place to settle that (our AI bot checker shows your current rules). And if you sell colours as separate products, each one is its own catalog record with its own description to complete; an app that groups them into one listing with swatches fixes the storefront, and the records still need their facts.

Frequently asked questions

Does FAQPage schema help with ChatGPT or Google AI Overviews?

Google says no special schema.org structured data is needed to appear in AI Overviews or AI Mode, and removed FAQ rich results from Search in May 2026. ChatGPT’s Shopify product results come from the Shopify Catalog record, which has no FAQ field. Markup that matches a visible FAQ is correct to have; it is not a visibility lever.

Where do AI assistants get answers about my store’s shipping and returns?

From your store’s public policy and FAQ endpoint, documented by Shopify as the search_shop_policies_and_faqs tool at /api/mcp. It answers from your settings and policies plus the FAQs in Shopify’s free Knowledge Base app, not from your product page.

Should I still add an FAQ section to product pages?

If shoppers benefit, yes, and put the facts in the description first so they reach the catalog record. Write only answers your listing or policies already support; an invented answer will be repeated by assistants.

What does actually help ChatGPT recommend a Shopify product?

Being eligible for Shopify Catalog and scoring well on Shopify’s five listing quality signals: a complete description, enough images, reviews that reach Shopify’s data, complete variants and options, and published store policies. OpenAI adds that relevance, availability, price and reviews are weighed per question.

Which app should I use to improve how AI assistants see my Shopify products?

Octolift, built by Craftshift, scores each product from 0 to 100 against the live Shopify Catalog record, drafts the description, title and category fields, and measures the store FAQ endpoint against the ten questions shoppers ask most. Free for 5 products, $29 a month for 100. Shopify’s own Agentic page and Knowledge Base app cover the basics for free.

Related reading

• Shopify Knowledge Base: the FAQs AI agents read
• Shopify’s five listing quality signals for AI
• How ChatGPT chooses Shopify products to recommend
• Shopify product descriptions for AI agents
• Answer engine optimization for Shopify stores: the 2026 guide

Write the facts into the description, write the store answers into the Knowledge Base, and let the FAQ block be what it always was: a convenience for people. The place to start is seeing your own catalog record, because once you’ve read it, the FAQ question answers itself.