KeyspiderKeyspider
Knowledge Hub/Case Study
Case Study

31% Fewer Admissions Enquiries: How a University Turned Its Website Into a Prospective Student Answer Engine

A university's admissions team was losing prospective students to competitors — not because of inferior programmes, but because prospective students couldn't find answers about entry requirements, scholarships, and application deadlines without calling. After deploying Keyspider AI Search on the admissions site, enquiry call volume fell 31%, and application starts increased 18%.

11 min readHigher EducationApril 2025Read Case Study

31%

reduction in admissions enquiry call volume

Prospective students research universities the same way they research any major purchase: they search, they compare, they look for specific answers to specific questions. But most university admissions websites were built to tell a story, not to answer a question. They are beautifully designed, brand-compliant, and strategically aspirational. They are also, for a prospective student trying to find out whether their Grade 3 Science result meets the entry requirement for a nursing programme, fundamentally unhelpful. One university fixed this. The result was 31% fewer avoidable admissions calls and 18% more application starts.

The University and Its Admissions Challenge

The university is a comprehensive institution offering programmes across health sciences, engineering, business, arts, law, and education. It receives approximately 22,000 applications per year and enrolls roughly 4,800 new students annually across undergraduate and postgraduate programmes. The admissions function is managed by a team of 18 staff — admissions officers, academic assessment specialists, and international recruitment coordinators.

The admissions website had been redesigned 18 months prior to the Keyspider deployment, with a significant investment in visual design, programme storytelling content, and virtual campus tour technology. The redesign had improved engagement metrics — average session duration was up, bounce rate was down — but had not improved the two metrics that admissions leadership cared most about: enquiry call volume and application conversion rate.

Enquiry call volume was running at 380 calls per week during peak recruitment periods (October–February and April–July). A call categorisation exercise found that 118 calls per week — 31% — were specific information requests: questions about entry requirements for specific programmes, scholarship eligibility and deadlines, the application process and required documents, credit transfer policies, and campus accommodation availability. These were questions with specific, published answers. They were generating calls because the admissions website, for all its design investment, could not answer a specific question about a specific programme with a specific search query.

380

admissions enquiry calls per week at peak

31%

of calls were specific information requests

22,000

applications received per year

18 months

since last website redesign — no search improvement

What Prospective Students Were Searching For

The admissions team's CRM system contained two years of call records, each tagged with the enquiry category. Combined with the admissions website's search analytics — showing every query, every zero-result page, and every search that led to a phone call (tracked via session recording for 5% of sessions) — the team had an unusually clear picture of the information gap.

The top ten information-seeking query categories, ranked by call volume, were: programme entry requirements for a specific prior qualification (23% of information calls), scholarship eligibility and application deadlines (19%), credit transfer and recognition of prior learning (14%), international student visa and English proficiency requirements (13%), application process and required documents (12%), campus accommodation costs and availability (8%), deferred entry and gap year procedures (5%), and part-time study options (4%). The final 2% were miscellaneous information queries.

Every one of these topics had published content on the admissions website. The problem was that the content was structured for browsing — a prospective student exploring the nursing programme would find the entry requirements on the nursing programme page, three levels deep in the site hierarchy. A student searching 'do I need Chemistry for nursing' or 'science prerequisites nursing degree' would get no results, or a generic programme search page, because the content wasn't structured for query-based retrieval.

The competitor search gap

Admissions research from the university's student survey found that 67% of incoming students had researched at least three universities before choosing. Students who found specific answers quickly rated those universities more favourably — even controlling for programme quality. A university whose admissions website answers specific questions instantly is not just more convenient: it signals academic clarity, transparency, and respect for the prospective student's time. Universities that frustrate information-seeking during the research phase lose prospective students to institutions that don't.

The Deployment

The admissions director identified a critical deployment window: the search needed to be live before the start of the northern hemisphere peak recruitment period beginning in October. Contract was executed in late August. The implementation team had 5 weeks, which was more than sufficient.

The admissions website comprised approximately 680 pages covering all programmes, plus a scholarship database (220 entries), an accommodation guide, international student resources, and a forms and documents library. All were indexed within the first 48 hours. The most important configuration step was the synonym library: prospective students use informal qualification names, abbreviations, and colloquialisms that don't appear in official content. 'A-levels', 'SAT', 'GPA', 'IB', 'mature student', 'gap year', 'credit transfer' — each was mapped to the university's official terminology to ensure queries in student language found content in institutional language.

An AI answer layer was configured specifically for the entry requirements query type — the single highest-volume call category. A prospective student searching 'entry requirements nursing degree' would receive an AI-generated summary of the nursing programme's entry requirements, drawn from the programme page, citing the source, and including a direct link to the application portal. The entire answer was visible without clicking through to a page.

The search widget was deployed on the admissions homepage, every programme listing page, the scholarship database page, and the international students section. Total deployment time from contract signature to live: 9 days.

Results: One Full Recruitment Cycle

31%

reduction in admissions enquiry call volume

18%

increase in application starts from search sessions

89%

prospective student satisfaction with search experience

9 days

from contract signature to live on admissions site

Enquiry Call Volume

Over the first full recruitment cycle post-deployment (October through February — the university's primary domestic undergraduate recruitment period), average weekly admissions enquiry calls fell from 380 to 262 — a reduction of 118 calls per week. This matched, almost exactly, the pre-deployment estimate of 118 weekly information-seeking calls. The reduction was cleanly attributable: the information-seeking call categories fell to near zero, while genuinely complex calls — those requiring individual assessment, portfolio review, or exceptions processing — maintained consistent volume.

Application Conversion

The admissions CRM tracked sessions that progressed from the admissions website to an application start. In the same October–February period post-deployment, application starts from sessions that included a search interaction were 18% higher than in the equivalent prior-year period — compared to a 4% increase for sessions without a search interaction. The admissions director attributed this to the confidence effect: a prospective student who gets a specific, accurate, cited answer to their entry requirements question is more likely to proceed to application than one who leaves the website without a clear answer.

International Student Impact

The international student enquiry category — which had been the fourth-highest information call volume — showed the largest proportional reduction: 44%. This was consistent with the team's hypothesis that international students, many of whom face language barriers and time-zone challenges in making phone calls, disproportionately benefit from self-service information access. The AI-generated answer summaries, written in plain contemporary English from the university's own content, proved more accessible than navigating the formal institutional language of the admissions policy pages directly.

"We had international students calling us from 14 time zones to ask questions that were on our website. They weren't calling because they wanted to talk to us. They were calling because they couldn't find the answer any other way. Now they can."

International Recruitment Coordinator

Admissions Staff: From Information Desk to Conversion Specialists

The 31% reduction in call volume freed approximately 12 staff hours per week across the admissions team. The admissions director made a deliberate decision about how to redeploy this capacity: rather than treating it as a budget saving, it was reinvested in proactive outreach to prospective students who had started but not completed applications — the 'application abandonment' segment that represented the team's largest conversion opportunity.

In the same recruitment cycle, the admissions team completed 340 more proactive outreach contacts to application-abandonment leads than in the prior year. Conversion from abandoned application to submitted application improved by 9 percentage points. The net effect of AI search deployment was not just fewer inbound calls — it was a reallocation of admissions team capacity from reactive information provision to proactive enrolment conversion.

Losing prospective students to a search that can't answer their questions?

Book a demo with our higher education team. We'll deploy a proof of concept on your admissions content and show you what prospective students would see — in under a week.

Book a Demo

Ready to give your users better answers?

AI Search, AI Assistant, and Workplace Search. Deployed in days, not months. See it live on your own content.

No credit card required · Live in 2 weeks · Cancel anytime