Higher education marketing is judged on applications, and it is decided by things most marketing plans never mention: whether a course page states its fee in text, whether one course has one URL, whether an answer engine is allowed to read the site at all.
In August 2026 we audited 64 UK universities, university colleges and specialist institutions against seven identical technical checks. We excluded the Russell Group and the largest civic universities, because their resourcing and their problems are different. What came back reads less like an engineering report and more like a marketing diagnosis.
This piece takes the findings and translates them into the terms a marketing director works in: cost per enquiry, share of voice on course terms, conversion from enquiry to application, and the credibility of the numbers you report at the end of the cycle.
The five numbers a marketing team should care about
Each of these is a technical measurement. Each one sits directly upstream of a marketing result.
Read the top line first. Ninety two percent of the readable robots.txt files name no AI crawler at all, in either direction. That is not a policy decision, it is an absence of one, which means nobody in the institution owns how the site appears in AI answers. In most cases marketing discovers the position only when a prospective student quotes a wrong fee back to an open day host.
The middle two lines are the recruitment story. Forty five percent of identified course pages keep fees out of server rendered HTML, and thirty four percent do the same with entry requirements. Those are the two questions that qualify or disqualify an applicant. When the page will not answer them, the applicant finishes the comparison somewhere else, on a page that holds your numbers secondhand.
Where marketing objectives actually get decided
The pattern across the sample is consistent: the objective sits with marketing, and the dependency sits with a different team, on a platform a third team owns. That gap is why plans that are right on paper still underdeliver.
None of the dependencies in that table are campaign work. They are template work, information architecture work, and data capture work, which is exactly why they fall between teams. Getting them done is a question of sequencing rather than budget, and that sequence is what higher education digital strategy has to commit to in writing.
A course is now compared in four places at once
Ten years ago a course competed for a search ranking. It now competes on four surfaces simultaneously, and each one reads the same page in a different way.
The useful part of this picture is the overlap. One set of page level fixes, facts in HTML, one canonical course URL, Course and FAQPage structured data, and a stated crawler policy, improves all four surfaces at once. That is unusually good leverage for a sector where marketing budgets are under review.
The aggregator card is the one that costs the most and gets discussed the least. League table and course comparison sites rank on your figures because they publish them in extractable form and you do not. Publishing course facts so answer engines can read them is how an institution becomes the source rather than the subject.
What this means for the marketing plan
Paid search
Course intent is the most expensive traffic a university buys and the most often wasted. When the click lands on a subject hub or a finder result rather than the course page, the intent that justified the bid is discarded. Thirteen percent of the sample split the course estate across more than one domain, which makes that misrouting structural rather than accidental. Fix routing and templates before the bid strategy, and see SEO for universities for the dependency order.
Organic and AI visibility
These are now one workstream, not two. The same missing facts that stop an answer engine quoting a fee also weaken the page against a ranking algorithm that rewards completeness. The one bright line in the audit is llms.txt, published by seven institutions in 64, which for a convention this young is ahead of the curve.
Brand and reputation
Brand campaigns run on the assumption that the site will hold up under the attention they create. In practice attention arrives on pages that were never designed for a cold visitor. The check is simple: take the three pages your brand campaign will drive traffic to, and confirm a stranger can act on each of them.
Reach and inclusion
Accessibility belongs in the marketing plan because an inaccessible page is an unreachable audience. WCAG 2.2 for university websites covers what changed, and accessibility statements covers the published commitment that goes with it.
Reporting you can defend
Long cycles, multiple devices, and consent-gated analytics make session-based reporting fragile in this sector. Capture campaign parameters and course context server side at submission, so every enquiry record carries its own provenance. Then report on what you can count honestly: qualified enquiries by course, response time, enquiry to application rate, and offer to acceptance.
University of East London: what the same fixes produced
UEL is the clearest illustration in our own work of marketing objectives resolved through platform and content work. We moved the estate from Sitecore to Drupal on Acquia, mapped journeys across five audiences, automated fee and deadline updates so course pages stayed current without manual edits, and unified the data layer so campaign, enquiry and course context arrived together.
The outputs we own are measurable: 90 percent WCAG 2.1 AA conformance on the measured estate, 46 percent fewer exits on key journeys, and 99.9 percent uptime through peak clearing periods. UEL reports a 73 percent rise in applications and a 32 percent rise in enrolments over the same period, alongside a 30 percent rise in relevant traffic. Those are their figures, on their hypothesis, and worth separating from ours.
The full engagement is written up in the University of East London case study.
A ninety day sequence
- Audit your ten highest value course pages for six facts in HTML: fee, entry requirements, duration, award, start date, mode of study.
- Establish one canonical URL per course, and de-index or consolidate duplicates and finder result pages.
- Point every course-intent paid keyword at that canonical page.
- Add Course and FAQPage structured data to the course template, not page by page.
- State an AI crawler position in robots.txt, and confirm the edge is not blocking crawlers you intended to allow.
- Capture campaign parameters and course code server side on every enquiry form.
- Set and publish a response time commitment for enquiries, measured weekly by course area.
Frequently asked questions
What is higher education marketing responsible for in 2026?
The same demand generation as before, plus the readability of the pages demand lands on. Channel work alone cannot compensate for a course page that will not answer the two qualifying questions an applicant has.
Why does technical work belong in a marketing plan?
Because the measurements that decide marketing performance are technical. Facts in HTML, one URL per course, structured data and crawler policy each map directly to a marketing objective, and none of them are campaign deliverables.
How do universities appear in AI answers?
By publishing extractable facts, allowing the crawlers you intend to allow at both robots.txt and the network edge, and structuring course data so it can be quoted. Ninety two percent of the institutions we audited have taken no position at all.
Is this audit representative of the UK sector?
It covers 64 institutions outside the Russell Group and the largest civic universities, audited in August 2026 against seven identical checks. Every percentage states its own denominator, and we never infer absence from a request that failed.
Where should a small marketing team start?
With one course page and one paid keyword. Fix the facts, fix the destination, measure the enquiry rate for a month. It is the shortest route to an internal case for doing the same across the estate.
Read next
The full dataset, method and per-check results are in the UK higher education AI discoverability report. Our higher education practice page sets out how strategy, design, engineering and marketing run as one accountable team, and student recruitment marketing, university website design and choosing a CMS for a university estate go deeper on the three areas this article touches.
Bring this dispatch into a working session - one page in, scoping memo out.
Brief Foyer