AI Search Is Already Citing Schools as Sources - Is Yours One of Them?

It's a reasonable assumption: rank well on Google, and AI search will naturally follow along behind it. The data doesn't support that assumption anymore. Research tracking which sites AI engines actually cite found that the overlap between top Google rankings and AI-cited sources has collapsed, from around 70 per cent down to under 20 per cent in a little over a year. Being findable and being citable have become two separate jobs, and most school websites are still only doing the first one.

This piece is a level deeper than “is my school visible in AI search” it's specifically about what makes a page worth quoting once an AI system finds it.

What actually earns a citation

Three moves show up repeatedly in research on what gets a page quoted inside an AI-generated answer, each lifting citation rates by roughly 30 to 40 per cent on their own.

Citing primary sources.

Pages that reference real, checkable sources - exam board data, ISI or Ofsted reports, your own results tables, get quoted more than pages making unsupported claims. If a claim can't be traced back to something concrete, an AI system has no way to verify it, and tends to skip it.

Attributed statistics.

“Outstanding pastoral care” is a claim. “94% of our Year 11 leavers progressed to their first-choice sixth form (2025 cohort)” is a citable fact, with a number, a context, and a date attached. AI systems overwhelmingly prefer the second kind of sentence - it's the difference between an opinion and a fact they can safely repeat.

Quoting named experts.

A line attributed to your actual Head of Admissions or Head Teacher, by name, carries more weight than the same sentence written anonymously as “the school.” Named expertise is one of the clearest trust signals AI systems are trained to look for.

The structural habits that help

Alongside what you say, how a page is built affects whether an AI system can use it at all. The pages that perform best tend to share a handful of habits: a short, plain-English summary near the top of longer pages rather than making a reader (or a crawler) dig for the point; claims stated clearly with the supporting evidence sitting right next to them, not three paragraphs away; a genuine Q&A section answering the specific questions parents actually ask, not a generic FAQ written for search engines a decade ago; data presented in simple tables where it makes sense - open day dates, fees, class sizes, results, rather than buried in prose; and visible authorship and a genuine “last updated” date, both of which act as freshness and trust signals.

One engine isn't the whole picture

Different AI systems cite very differently from one another. Perplexity typically references around twice as many sources in a single answer as ChatGPT does, and the two tools share only a small fraction of the domains they actually cite. Being picked up by one doesn't mean you're picked up by another - this needs to be a deliberate, ongoing check across systems, not a single box to tick.

A quick technical check worth making

One surprisingly common and easily missed issue: older websites sometimes carry a broad, historic robots.txt rule that was written long before AI crawlers existed, and it can end up silently blocking GPTBot, Google-Extended or PerplexityBot from accessing the site at all, meaning no matter how good the content is, it's technically invisible to the systems it's meant to be reaching. Worth having someone check this specifically rather than assuming it's fine.

If you'd like us to check whether your site is technically blocking AI crawlers, or review how citable your current content actually is book a discovery call and we'll walk through it together.

Sources referenced for this piece: research on Generative Engine Optimization (GEO) and AI citation patterns summarised via TechTimes, Enrich Labs, and Kick Ads' 2026 GEO guides.