AI Visibility Audit
This report is confidential. Enter the access code provided by Saigon Digital to continue.
Incorrect code. Please try again.
Prepared by saigon.digital · September 2026
This report is confidential. Enter the access code provided by Saigon Digital to continue.
Families increasingly ask ChatGPT, Perplexity and Google's AI for a shortlist before they ever reach a school website. This audit covers what those engines can currently read about UWC Atlantic, what is blocking them, and the three fixes that matter most.
Saigon Digital helped us increase awareness of BVIS and support admissions for the opening of our new campus. Through a focused SEO and AI visibility strategy, they helped more parents and families discover BVIS when searching online for international school options. Their work strengthened our digital presence and helped us reach prospective families at an important time for the school.
See how we've helped brands grow. Read our case studies and learn more about what we do.
View Case StudiesA snapshot of where UWC Atlantic stands in the AI search era, and the opportunity cost of the current gap.
The good news first. UWC Atlantic's website is fully open to the AI engines, which is not true of every school we look at. The problem is what those engines find once they arrive.
The headline finding. The file that tells every search engine and AI crawler where to find your content,
uwcatlantic.org/robots.txt, currently points them at the sitemap of a car finance company.
Not a typo in a page, a live instruction on your domain. Your own sitemap of 192 pages is never mentioned to them at all.
Taken together, the three findings below mean the AI engines are welcome to read UWC Atlantic, are given the wrong directions on arrival, and then find nothing machine readable that identifies the school, its location, its IB programme or its leadership. All three are configuration fixes rather than rebuilds.
Three issues and one piece of good news. Every item was checked live on 21 September 2026 and each one is verifiable from a browser in under a minute.
The whole file currently reads:
User-agent: *
Disallow:
Sitemap: https://www.motonovofinance.com/sitemap.xmlMotoNovo Finance is a Cardiff based car finance lender. It is not connected to the school. The most likely explanation is a template reused by a developer who had worked on both sites.
Two consequences follow. First, every engine that reads this file is handed another company's domain as the authoritative
index of your content. Second, that URL returns a 403 Forbidden, so the engines are pointed at a page they
cannot even open. Meanwhile your own sitemap at uwcatlantic.org/sitemap.xml works perfectly and lists
192 pages, and nothing anywhere tells a crawler it exists.
Structured data is the machine readable summary a page carries about itself: what the organisation is, where it is, what it offers, who leads it. It is the single most reliable way an AI engine confirms a fact well enough to repeat it.
UWC Atlantic's homepage contains no structured data blocks at all. An engine wanting to state that UWC Atlantic is an IB World School in the Vale of Glamorgan has to infer it from page copy, or fall back on third party sources such as Wikipedia and directory listings, which is where the outdated entries tend to live.
All three of the schools we compared do publish it. That comparison is in the next section.
Your homepage meta description, the sentence engines and search results quote when they summarise a site, currently reads in full:
"Atlantic College UWC Principal St Donats Naheed Bardai"
That is a keyword list rather than a description, and the same string is duplicated into the Open Graph and Twitter descriptions, so it is also what appears when anyone shares a link to the school on social platforms or in a messaging app. A model asked to describe UWC Atlantic in one line has nothing usable to draw on here.
We requested the homepage as each of the major AI crawlers in turn. GPTBot, PerplexityBot, ClaudeBot and Google-Extended were all served the page normally. Many schools we audit block these outright, often without realising it, and then wonder why they are absent from AI answers.
UWC Atlantic does not have that problem, which means the three fixes above are additive. You are not digging out of a hole, you are supplying content to engines that are already willing to read it.
Tested live on 21 September 2026 by requesting uwcatlantic.org as each crawler's own user agent.
Free to crawl the site. Finds no structured data and a keyword string where the description should be.
Free to crawl. Perplexity leans heavily on citable structured sources, which the site does not currently provide.
Free to crawl, and directed by robots.txt to a car finance sitemap that returns 403.
Governs use in Gemini and AI Overviews. Allowed in, with no schema to confirm any fact about the school.
A note on method. We have deliberately not published claims about what any engine says about UWC Atlantic today. Model answers vary by user, by country and by week, and a test run outside the UK would not reflect what a family in Britain sees. What this section reports is what the engines are able to read, which is measurable, stable, and the part you control. If it is useful, we are happy to run live prompts together on a call so you see the answers yourself rather than take ours on trust.
The three schools named in our first email, checked the same way on the same day.
This is the gap worth closing first, and it is worth being clear about why. It does not mean those schools outrank UWC Atlantic in AI answers, and we are not going to claim that without showing you the evidence live.
What it does mean is that when an engine needs to confirm a fact about each school, three of the four hand it a clean answer and one asks it to guess. For a school whose distinctiveness is the whole proposition, an IB World School in a twelfth century castle with a global scholarship intake, leaving engines to infer that from prose is a real cost.
None of these is a rebuild. The first one your developer can do this week.
Replace the MotoNovo line with your own sitemap so crawlers are pointed at all 192 of your pages:
User-agent: *
Disallow:
Sitemap: https://www.uwcatlantic.org/sitemap.xmlWe would do this one regardless of whether we work together. It costs nothing and it is currently sending your crawl budget to another company.
A single block on the homepage stating the legal name, the St Donats address, the charity number, the IB World School status, the Principal and the admissions contact. Extended onto the admissions and programme pages it gives engines a confirmed set of facts to repeat rather than infer, and it is what the other three schools already have.
One clear sentence describing the school, replacing the keyword string, and carried properly into the Open Graph and Twitter tags so shared links read correctly too.
Alongside it, uwcatlantic.org/llms.txt currently returns a 404. That file is becoming the conventional way to give
AI engines a plain summary of an organisation and point them at the pages that matter. It is early, it is cheap, and very few schools have one yet.
As promised, no cost and no pitch attached to this report.
Nick Rowe, CEO & Co-Founder, Saigon Digital