AI visibility ranking of Spanish universities
The observatory of which universities the AI recommends when someone asks where to study — measured across ChatGPT, Gemini, Claude, Perplexity and updated every month.
Search has changed hands
For twenty years, getting informed meant searching on Google: you typed, ten blue links appeared and you chose. Today, more and more people don’t search, they ask — they talk to ChatGPT, Gemini or Perplexity the way you’d consult an expert — and the AI doesn’t return ten options: it returns one answer with three or four names.
For institutions it’s a seismic shift. On Google you could sit in eighth place and still get clicks; in an AI answer, if you’re not in that shortlist, you simply don’t exist for whoever is asking. Optimizing to show up there has a name: GEO (Generative Engine Optimization), the successor to SEO. But before you optimize you have to measure: this analyzes 52 institutions across 4 AI models, and it’s updated every month.
AI Visibility Index
One number per university: the AI Visibility Index — how much the AI recommends it today. The columns on the right break that visibility down model by model, from a broad study across student topics — degrees, master's, MBA, online, bootcamps — with 60+ real queries to the models. Nobody reaches 100 — not even the leader, so you can see how much room there still is, and how steeply it drops right after the top.
| # | University | AI Index | ||||
|---|---|---|---|---|---|---|
| 1 | 66 | 41 | 43 | 33 | 12 | |
| 2 | 57 | 42 | 22 | 28 | 31 | |
| 3 | 56 | 43 | 38 | 31 | 27 | |
| 4 | 36 | 26 | 13 | 15 | 16 | |
| 5 | 36 | 41 | 47 | 35 | 30 | |
| 6 | 33 | 24 | 20 | 11 | 8 | |
| 7 | 29 | 18 | 19 | 14 | 10 | |
| 8 | 26 | 27 | 32 | 25 | 33 | |
| 9 | 25 | 21 | 17 | 13 | 18 | |
| 10 | 24 | 21 | 15 | 13 | 15 | |
| 11 | 22 | 31 | 25 | 18 | 29 | |
| 12 | 21 | 13 | 12 | 6 | 3 | |
| 13 | 19 | 21 | 20 | 22 | 14 | |
| 14 | 18 | 14 | 19 | 8 | 17 | |
| 15 | 18 | 18 | 15 | 10 | 11 | |
| 16 | 18 | 21 | 19 | 14 | 15 | |
| 17 | 17 | 13 | 5 | 8 | 5 | |
| 18 | 14 | 8 | 2 | 5 | 2 | |
| 19 | 14 | 11 | 7 | 5 | 2 | |
| 20 | 14 | 13 | 19 | 5 | 7 | |
| 21 | 14 | 13 | 12 | 19 | 21 | |
| 22 | 14 | 14 | 19 | 17 | 19 | |
| 23 | 13 | 7 | 6 | 2 | 2 | |
| 24 | 12 | 10 | 14 | 8 | 11 | |
| 25 | 11 | 10 | 6 | 7 | 4 | |
| 26 | 11 | 11 | 12 | 8 | 13 | |
| 27 | 11 | 13 | 18 | 18 | 13 | |
| 28 | 10 | 6 | 4 | 2 | 2 | |
| 29 | 10 | 11 | 10 | 10 | 12 | |
| 30 | 9 | 7 | 2 | 2 | 2 | |
| 31 | 9 | 5 | 2 | 2 | 2 | |
| 32 | 9 | 5 | 12 | 2 | 4 | |
| 33 | 8 | 9 | 9 | 9 | 12 | |
| 34 | 8 | 11 | 12 | 14 | 14 | |
| 35 | 7 | 4 | 19 | 7 | 7 | |
| 36 | 6 | 3 | 2 | 2 | 2 | |
| 37 | 6 | 3 | 2 | 2 | 2 | |
| 38 | 6 | 3 | 11 | 5 | 2 | |
| 39 | 6 | 5 | 13 | 5 | 6 | |
| 40 | 5 | 3 | 4 | 2 | 5 | |
| 41 | 5 | 3 | 8 | 2 | 7 | |
| 42 | 5 | 5 | 17 | 2 | 5 | |
| 43 | 5 | 6 | 12 | 10 | 5 | |
| 44 | 4 | 2 | 2 | 2 | 2 | |
| 45 | 3 | 4 | 3 | 4 | 4 | |
| 46 | 2 | 2 | 2 | 2 | 3 | |
| 47 | 2 | 2 | 2 | 2 | 8 | |
| 48 | 2 | 2 | 2 | 2 | 2 | |
| 49 | 2 | 3 | 4 | 2 | 4 | |
| 50 | 2 | 2 | 2 | 2 | 2 | |
| 51 | 2 | 4 | 4 | 2 | 2 | |
| 52 | 2 | 2 | 2 | 2 | 3 |
Compare two universities
Pick two institutions and pit them against each other in detail, model by model.
Who leads each field
The first thing every institution wants to know: how it’s doing in its own field.
Complutense
U. Europea
U. de Granada
UNIR
VIU
UOC
IE University
ESADE
OBS
Ironhack
DKS
Nuclio
How the Index is calculated
The Index combines five weighted signals, all measured with real data on how the models respond. The result is an absolute number from 0 to 100 — not a ranking where first place scores 100 «just because».
Each AI recommends differently
There isn’t one AI: there are four, each with its own sources and biases. Being on top in Gemini doesn’t guarantee showing up in ChatGPT. Knowing which model your problem is in is what separates a headline from a useful diagnosis.
Its favorites in the study: Universitat de Barcelona, Navarra and Carlos III.
Closely tied to classic SEO. Its top picks: UB, Universidad Europea, Complutense and Deusto.
More conservative; leans on track record. Its top picks: UB, Complutense and Carlos III.
Rewards verifiable data. Its top picks: Carlos III, Navarra, UB and ESADE.
How students search
Depending on what the student is looking for, the AI spreads names wider or concentrates them more — and what makes you win changes.
The AI hands out more names. It's your best way into the conversation.
The AI picks a winner. Third-party reviews and attributes decide it.
Historic, authoritative brands win. Hard to shift in the short term.
Whoever publishes prices and terms clearly and in a structured way wins.
Placement data, partnerships and a LinkedIn presence all carry weight.
How the AI treats you
It’s not the same for the AI to mention you in passing as to recommend you, and each institution carries certain attributes in the model’s «mind». We show a selection of institutions.
Out of every time the AI names it: how many times does it truly recommend it, and how many does it just mention it in passing?
The words the AI uses to describe each institution. Green = in favor · red = against.
Where the AI gets its information
This is the most actionable GEO metric of all: which sources the AI cites when it talks about universities. It leans on education portals and comparison sites, media rankings (El Mundo, CYD) and the universities' own pages — that’s where you need to be, not just on your own website.
What people ask the AI
The sector’s highest-volume queries on Google (Spain). When the AI answers these, the question is whether you show up.
| Query | Searches/mo |
|---|---|
| “best universities in Spain” | 2,400 |
| “distance-learning university in Spain” | 1,300 |
| “private universities in Spain” | 590 |
| “online master's in education” | 590 |
| “study psychology online” | 390 |
| “online master's in HR” | 390 |
How to improve your AI visibility
AI visibility responds to concrete signals. These are the ones that weigh most:
Pages with data, comparisons and FAQs the AI can extract and cite.
The sources models treat as trustworthy by default.
Reddit, student forums and reviews shape how the AI describes you.
Program listings marked up so the AI understands them without ambiguity.
Classic SEO still counts: it feeds search-linked models like Gemini.
A consistent name, campuses and data across the site keep the AI from confusing you.
AI visibility will be decided over the next two academic years
Traditional search engines took years to lock in their winners; the AI is doing it in months, and positions are still moving. Whoever understands now how that visibility is won will hold an advantage that’s hard to claw back once the market matures. And we’ll say it again because it matters: the drop-off right after the leader is brutal. There’s room for everyone — and the first step is always the same, knowing where you stand.
See how your university shows up
Volcanz’s Radiografía analyzes your domain for free: your AI visibility Index, which models you appear in, who’s overtaking you and where to start your GEO.
Get my free RadiografíaFrom diagnosis to action, in 4 steps
Can a university improve its Index? How?
Yes, and that’s the interesting part: almost nobody has started, so the room to grow is enormous. You work on citable, structured content (comparisons, FAQs, employability data), a presence on Wikipedia and in reference media, reviews on forums, and a solid technical and SEO foundation. It takes months, but whoever starts today gains an advantage that’s hard to claw back.
Does showing up in AI actually bring in students?
More and more. Choosing where to study is one of the decisions people are delegating most to AI — comparing prices, programs and career outcomes is exactly what a model does well. If the AI doesn’t name you in that discovery and comparison phase, you disappear from the process before the student even visits your website.
Can you pay to show up in ChatGPT, like Google Ads?
No. Today there are no «ads» in the answers: the AI names you if it considers you a relevant, reliable source. That’s why GEO is so valuable — it’s earned with authority and content, not by buying the spot, and that makes it far more defensible in the long run.
Does the AI make up universities or give outdated data?
Sometimes it does: it mixes up campuses, attributes degrees that don’t exist or quotes old prices. If your official information isn’t clear and consistent, the model fills the gaps with third-party sources — and that’s where you lose control of the narrative. Looking after how the AI «reads» you is part of GEO.
Does GEO replace SEO or do they complement each other?
They complement each other. SEO still brings in traffic and, on top of that, feeds search-linked models like Gemini, which is closely tied to Google. GEO adds the new layer: getting the AI to recommend you, not just to have you appear in a list of ten links.
How long until an improvement shows?
It depends on the signal: structured data and citable content can show up in weeks; authority (Wikipedia, media, links) is slower, taking several months. That’s why we measure every month: to see the movement and not fly blind.
How is the Index calculated?
In two layers. The per-model columnscome from a study across the main student topics (degrees, master's, MBA, online, bootcamps…) — 60+ queries to ChatGPT, Gemini, Claude, Perplexity with web search in the Spanish market — scoring how often each institution appears and how prominently. The Indexthen blends that with the broader volume of AI mentions and authority signals (Wikipedia, media, links). It's an absolute score: 100 would mean being recommended everywhere, in every model — so nobody reaches it, and the room to grow is real.
Why isn’t my university in the ranking?
We show a sample of 52 institutions. If yours isn’t here, measure it instantly with the free Radiografía.
How often is the observatory updated?
Monthly. AI visibility changes month to month.