How AI assistants decide which stores to mention

Each assistant searches a different index

An index is the library of pages a search system has stored; a page outside it cannot be found. What each company says:

The takeaway: being indexed by Google does not mean ChatGPT, Perplexity or Claude can read you. If one company's crawler is blocked, that assistant will rarely cite you. OpenAI says it plainly: sites that opt out of OAI-SearchBot will not be shown in ChatGPT search answers.

One question becomes several narrower searches

Assistants rarely search with the shopper's exact words. Google calls its method "query fan-out": AI Mode breaks a question into subtopics and issues many searches at once. Google's own example: for "how to fix a lawn that's full of weeds," the fan-out searches might include "best herbicides for lawns," "remove weeds without chemicals" and "how to prevent weeds in lawn."

OpenAI describes the same pattern in its help center. ChatGPT search typically rewrites your question into one or more targeted queries, and after the first results it may send more specific ones. Its example starts with "CCR8 immunotherapy drug development 2025" and narrows to "CHS-114 conference 2025." Bing's AI Performance report even has a column for it, "Grounding queries": the phrases the AI used when it retrieved content.

For a store, it might look like this (illustration of a likely breakdown, not a logged AI session):

• Shopper asks: best portable power station for van life under $500
• Possible searches: power station capacity for van life; LiFePO4 vs lithium ion cycle life; power station under $500 specs

The sub-searches are narrower than the original question. Our inference, not any vendor's words: a page built to answer one specific question has a better chance of matching one of them than a page that covers everything a little. What Google does say is that AI Overviews and AI Mode show a wider, more diverse set of links than classic search, which gives "more types of sites" a chance to appear. What that means for a small store is in Long-tail questions.

From the pages it can read, it quotes the passage that answers

Once it has results, the assistant opens pages, reads the text and picks a passage. A page has to clear three bars:

1. Reachable. The crawler is not blocked by robots.txt, a CDN or a firewall. Google's documentation specifically tells site owners to check CDN and hosting settings.
2. Indexed. Google says a page must be indexed and eligible to show with a snippet before AI Overviews can use it, and that important content should be available as text, not only in images.
3. On point. Perplexity says its system treats individual sections of a page as units in their own right. Bing's guidelines say a URL focused on one primary topic is more likely to be selected for grounding, with key information near the top.

Passages that answer first, state specs and prices in text and source their numbers are the easiest to lift. A hero image and a tagline give the assistant nothing to quote. To check the first two bars, see Can AI crawlers reach your store?; for the third, How to write a page AI will quote.

Third-party sources make up most of what gets cited

A supplement brand owner put it this way on the Shopify Community forum: strong on-page SEO, yet invisible in AI Overviews. "Is it possible that AI models are prioritizing third-party reviews and medical journals over direct merchant sites?"

Broadly, yes. Public research finds that most sources cited in AI answers are third-party sites, and brand sites are a small share. Google's own guide says its AI features can show what is said about products across the web, including blogs, videos and forum discussions. Review sites, Reddit threads, YouTube videos and industry directories are where assistants often look.

What works is the honest version: real customer reviews, answers to questions people already ask under your real name, and paid partnerships labeled as ads. Google adds that seeking inauthentic mentions is less helpful than it looks. We do not do sock puppets or seeded threads; see what we refuse to do.

Shopify stores have a second path: product data

Shopify's agentic storefronts let eligible stores' products be discovered and bought in ChatGPT, Google AI Mode and Gemini, Microsoft Copilot and Meta. Shopify's help center says it is active by default for eligible stores. You can review it in your admin under Sales channels > Agentic.

• ChatGPT and Copilot get your products through Shopify Catalog: title, description, options, images, price, availability and other attributes. Google AI Mode and Gemini get them through the Google & YouTube sales channel.
• ChatGPT acts as a discovery-focused referrer: the shopper checks out on your own store.
• Shopify notes that AI crawlers might also reach your store through the open web, and recommends complete product data in Catalog plus standard SEO practice.

Catalog decides whether your product card shows up in AI shopping results with the right price and stock. When a shopper asks something specific, like whether a unit can run a CPAP overnight, the assistant still answers from web pages, which means answer pages on your site. WooCommerce and custom stores lack this Shopify channel; for Google's AI features, Google points merchants to Merchant Center feeds.

What you can control

• Crawler access: robots.txt and CDN settings. Days to fix, and you can retest the same day.
• Indexing: submit your sitemap in Google Search Console and Bing Webmaster Tools, and link to important pages from your own site.
• One question, one answer page: pick specific questions your buyers really ask and answer each on its own page. It usually takes weeks to see those pages indexed and showing up in AI answers.
• Real off-site discussion: take part under your real identity, without astroturfing.
• Weekly checks: ask the same questions every week and record who gets cited; see how to check whether AI recommends your store.
• Shopify stores: open Sales channels > Agentic and fill in titles, descriptions and specs properly.

What you cannot control: which sources the assistant picks on a given day. Ask the same question twice and the citations can change, so no one can promise you will be recommended. Our approach is on the method page.

Common questions

Sources

1. OpenAI: Overview of OpenAI Crawlers, checked 2026-10-08
2. OpenAI Help Center: Searching the web with ChatGPT, checked 2026-10-08
3. Perplexity Crawlers, checked 2026-10-08
4. Perplexity: Architecting and Evaluating an AI-First Search API (200 billion URLs, authoritative domains, sections as units), checked 2026-10-08
5. Google Search Central: AI features and your website, checked 2026-10-08
6. Google Search Central: Optimizing your website for generative AI features on Google Search (fan-out example, mentions, Merchant Center), checked 2026-10-08
7. Google blog: Expanding AI Overviews and introducing AI Mode (2025-03-05), checked 2026-10-08
8. Google blog: AI Mode in Google Search: Updates from Google I/O 2025 (2025-05-20), checked 2026-10-08
9. Google: Google's common crawlers (Google-Extended), checked 2026-10-08
10. Bing Webmaster Guidelines, checked 2026-10-08
11. Bing Webmaster Blog: Introducing AI Performance in Bing Webmaster Tools Public Preview (2026-02-10), checked 2026-10-08
12. Claude Help Center: Does Anthropic crawl data from the web, and how can site owners block the crawler?, checked 2026-10-08
13. Shopify Help Center: Shopify agentic storefronts, checked 2026-10-08
14. Shopify Help Center: Shopify Catalog and product discovery for agentic storefronts, checked 2026-10-08
15. Shopify Community: Invisible in AI Overviews (store owner's question, 2026-07-14), checked 2026-10-08