You've done the work. Your schema markup is live, Google Search Console shows rich results firing correctly, and your product listings or FAQ panels are appearing in the SERPs. So why is it that when someone asks ChatGPT or Perplexity about your product category, your brand is nowhere to be found?
This is one of the most common frustrations we hear from e-commerce brands right now. And the honest answer is: Google rich results and AI search visibility are two fundamentally different things, built on different signals, serving different systems. Passing one does not mean you pass the other.
Google and AI Search Engines Are Reading Your Site Differently
Google's rich results system is relatively well-defined. You implement a supported schema type, Google's crawler picks it up, validates it against its guidelines, and if everything checks out, you get the enhanced SERP feature. The feedback loop is tight and the rules are published.
AI search engines like ChatGPT (via its browse or plugin features), Perplexity, and Gemini work differently. They are not simply running a schema validator. They are building a probabilistic model of which sources are trustworthy, authoritative, and worth citing when a user asks a question. Schema markup is one input into that model, but it is a long way from being the only one.
Think of it this way. Google rich results are a pass/fail technical check. AI citation is more like a reputation score built from dozens of signals over time. You can ace the technical check and still score poorly on reputation.
The Schema Types That Impress Google Do Not Always Help AI
Google has a relatively short list of schema types it uses to power rich results: Product, FAQ, HowTo, Review, BreadcrumbList, and a handful of others. These are well-supported and well-documented.
But AI engines are not constrained to that list. They can read and interpret a much wider range of schema types, and many of the most useful ones for AI visibility are types that Google largely ignores for rich results purposes.
For example:
- Organization schema with detailed properties like
knowsAbout,hasCredential, andsameAshelps AI engines understand who you are and whether you are a legitimate entity. - DefinedTermSet and DefinedTerm schema signal to AI that your content is definitional and authoritative on a topic. This is extremely useful for getting cited in explanatory answers.
- ClaimReview schema tells AI engines that your content has been fact-checked, which builds trust signals around your brand.
- SpeakableSpecification flags specific passages of your content as being particularly important for AI to read and surface.
None of these produce Google rich results. Most brands implementing schema for SEO purposes never touch them. But for AI visibility, they are often more valuable than the schema types that do produce rich results.
If you want to understand how some of these lesser-known types actually work, our post on using DefinedTermSet schema to get cited in AI glossaries goes into detail on implementation.
AI Engines Look at Your Entire Presence, Not Just Your Markup
This is the part that surprises most people. AI search engines are not just reading your website. They are drawing on training data, live crawls, third-party mentions, forum discussions, review platforms, and more. Your schema markup is one signal among many, and it only covers one source: your own site.
If your brand is not mentioned on authoritative third-party sources, AI engines have very little external validation to work with. They are cautious about citing sources they only know from the source's own self-description. That is roughly equivalent to trusting someone's CV that has no references.
Some of the off-site signals that matter include:
- Mentions in industry publications and news outlets
- Discussions on Reddit, Quora, and other forums that AI engines treat as genuine user sentiment
- Reviews on platforms like Trustpilot, Google, and Capterra
- Backlinks from editorially independent sources
- A Wikipedia or Wikidata presence (which acts as a strong entity anchor)
Rich results say nothing about any of this. You can have perfect schema markup and zero third-party presence, and AI engines will largely overlook you in favour of a competitor with messier markup but stronger external signals.
Entity Recognition Is the Missing Piece
AI search engines think in entities, not just keywords or schema types. An entity is a clearly defined, distinct thing in the world: a business, a person, a product, a place. For AI to cite you confidently, it needs to recognise your brand as a known, validated entity.
Entity recognition depends on several things working together:
Consistent NAP data across the web
Your name, address, and phone number should be identical across your website, Google Business Profile, directory listings, and any other platforms. Inconsistency confuses entity resolution.
sameAs links in your Organization schema
Your schema should include sameAs properties linking to your profiles on Wikipedia, Wikidata, LinkedIn, Companies House, Crunchbase, and any other authoritative directories. This tells AI engines that all of these profiles refer to the same real-world entity.
A Wikidata entry
You do not necessarily need a full Wikipedia article, but a Wikidata entry for your business is very achievable for most brands and significantly strengthens entity recognition. It acts as a structured data anchor that AI engines trust.
Google does not care much about any of this for rich results purposes. But AI engines care a great deal. This is why a brand can pass every Google schema validation check and still be invisible to AI.
Your Content Structure May Be Clear to Google but Opaque to AI
Google's indexing is largely keyword-matching and link-graph based. AI engines are trying to understand meaning, context, and authority. The same page can serve Google well and confuse an AI engine if the content is not structured in a way that makes your expertise obvious.
Some specific issues we see regularly:
No clear topical positioning
If your site covers too many loosely related topics without clear thematic depth in any of them, AI engines struggle to understand what you are genuinely an authority on. Google might still rank you for individual keywords. AI engines want to know: what is this brand the expert on?
Thin or generic content
AI engines are trained on vast amounts of text and are very good at identifying generic, surface-level content. If your product descriptions or blog posts could have been written by anyone about anything, you are not giving AI a reason to cite you over a competitor with more specific, detailed, original content.
Missing author and brand signals
Google does not require you to have named authors on every page. AI engines, particularly in the context of E-E-A-T signals, weight content more heavily when it is clearly attributed to a real person or organisation with verifiable credentials. Anonymous or generic content is harder for AI to trust.
Our post on what E-E-A-T is and why it matters for AI search explains this in more depth if you want to understand how AI engines assess credibility.
AI Crawlers May Not Even Be Reaching Your Pages
Here is a technical issue that catches a lot of brands off guard. If your robots.txt file is blocking AI crawlers, none of the above matters because these systems are not reading your content at all. And blocking AI crawlers by accident is surprisingly common, especially on older Shopify and WordPress setups where robots.txt has been modified over time.
Common culprits include blanket Disallow: / rules that were intended for specific bots but inadvertently catch AI crawlers, or plugin-generated rules that block anything not on a short whitelist. GPTBot (used by OpenAI), ClaudeBot (used by Anthropic), and PerplexityBot all need to be allowed if you want those systems to read your site.
Check your robots.txt file at yourdomain.com/robots.txt and look for any rules that could be blocking these agents. You can also review your server logs to see which AI crawlers are actually visiting and whether any requests are being denied.
We cover this in detail in our post on why your robots.txt might be blocking AI crawlers without you realising.
What to Actually Do About It
If your rich results are working but AI search is not citing you, the fix is not to tweak your existing schema. The fix is to build the broader AI visibility foundation that your current setup is missing.
Start with an honest audit of where you stand:
- Is your Organization schema complete, with
sameAs,knowsAbout, anddescriptionproperties? - Do you have a Wikidata entry?
- Are AI crawlers allowed in your robots.txt?
- Does your site have content with clear, specific topical depth, or is it mostly generic?
- Are you mentioned on any third-party authoritative sources?
- Do your reviews on external platforms reflect positively on your brand?
Each of these is a separate workstream, and most brands have gaps in several of them. The good news is that structured, methodical progress on these signals does produce results. It typically takes three to six months to see meaningful improvement in AI citation rates, but the compounding effect over time is significant.
At FlinnSchema, this is exactly what we work on with e-commerce brands: not just schema markup in isolation, but the full picture of what makes an AI engine trust and cite a business. If you are not sure where your biggest gaps are, a free AI visibility audit is a good place to start.
Frequently Asked Questions
Does fixing my schema markup guarantee AI search will mention me?
No. Schema markup is one factor among many. AI engines also weigh third-party mentions, content quality, entity recognition, and crawl access. Schema is necessary but not sufficient on its own.
Which AI crawlers should I make sure are not blocked?
The main ones to check for are GPTBot and ChatGPT-User (OpenAI), ClaudeBot (Anthropic), PerplexityBot (Perplexity), and Googlebot (which also feeds Gemini). Review your robots.txt and server logs to confirm none of these are being denied access.
How long does it take to start appearing in AI search answers?
Most brands see meaningful improvement within three to six months of implementing a proper AI visibility strategy. The timeline depends on how many gaps you are starting with and how quickly you can build third-party presence alongside your on-site work.
My competitor has worse schema than me but gets cited by ChatGPT. Why?
Almost certainly because they have stronger off-site signals: more third-party mentions, better entity recognition, stronger review presence, or a longer established reputation in their category. AI engines are not just reading structured data. They are drawing on a much broader picture of who is considered authoritative in a space.