The battle for visibility is shifting from search results to AI answers. How does your university perform?
Many of us have stopped clicking through to websites for quick answers. We read the summary at the top of the search results, or ask an AI tool, and move on.
Prospective students are no different. Nearly half (48%) of students who use AI have used it to explore their university options, according to UCAS.They’re asking questions like: AI which universities are good for psychology, what grades they’ll need, and what first year is actually like. More and more often, they get an answer before they ever reach a university website.
Our new report, How universities can win in AI search, looks at what this shift means for university web teams. Here’s a taste of what we found.
A new gatekeeper
For the last two decades, the goal was to get your page to the top of Google. That still matters, but ranking well no longer guarantees that anyone will visit. In the UK, almost seven in ten Google searches now end without a click, because the answer is already there on the results page.
So the aim has shifted. If your university isn’t part of that answer, a league table or another university will be instead.
This change is easy to miss, because much of it happens out of sight of your analytics. The University of Edinburgh saw AI referrals to its prospective student site grow from around 200 to around 1,300 in five months. Even those figures only count the people who clicked through. Many more will have had their question answered without ever reaching the site.
Students still trust your website
Students aren’t taking AI at its word, though. According to UCAS, 73% of those who use it say it has given them incorrect information, and 43% of students researching universities with AI still go to university websites.
AI tends to lead the early stages, while students explore options and build a shortlist. As they get closer to a decision, they turn to university websites to confirm what they’ve heard and find out more.
The job now is making sure AI points them there in the first place, and that your site gives them a clear answer when they arrive.
Simple fixes
There’s no shortage of new tactics promising to win at AI search. In our experience, the changes that make the biggest difference are fundamentals that good university websites have always got right.
Edinburgh’s story is a good example. Its entry requirements sat behind dropdown menus that AI tools couldn’t read. So the team added a plain companion page for each degree, generated automatically, with no extra work for editors. ChatGPT went from getting the grades right in 3 out of 11 searches to 9 out of 11. A simple change, but highly effective.
We saw something similar when we analysed AI answers for the University of Oxford, whose new website we designed and built. By structuring sections of course pages around key questions like ‘Why choose Oxford’ they were able to be mentioned in 93% of AI answers across high value queries like ‘Top universities for Computer Science’ or ‘Top universities for Climate change research’. This was far ahead of relevant competitors like the University of Cambridge who were only mentioned in 71% of the tested queries .
Governance is key
University sites are huge, publishing is often spread across hundreds of editors, and keeping thousands of pages accurate is a massive challenge.
Some universities are already acting on this. The University of Sussex, for example, has cut the number of people who can edit its public site from almost 1,200 to about 220. Fewer pages and fewer editors make it easier to stay on top of content and prevent the site from growing out of control.
The full report
This article only scratches the surface. The full report goes into the practical detail, from the six things findable content gets right, to which AI crawlers you should let in. It shows how to map the questions that should lead to your university, how to measure how often AI mentions you, and five questions to put to your web team.
