---
title: "Enrollment Management in Ireland: Funnel, Metrics and AI Agents"
description: "Enrollment management for Irish higher education: what it covers, the CAO funnel and its metrics, a worked example and where AI agents help admissions teams."
url: "https://www.skolbot.ai/en-IE/blog/ai-in-enrollment-management-admissions-teams"
language: "en"
dateModified: "2026-11-19"
---

# Enrollment Management in Ireland: Funnel, Metrics and AI Agents

Updated on November 19, 2026

Enrollment management for Irish higher education: what it covers, the CAO funnel and its metrics, a worked example and where AI agents help admissions teams.

## What is enrollment management?

Enrollment management, or enrolment management in Irish spelling, is the coordinated work a university, technological university or private college does to reach its student number targets: recruiting prospects, handling applications and offers, converting offers into registered students, and keeping them on course. In Ireland the calendar set by the [Central Applications Office](https://www.cao.ie/) (CAO) and the Leaving Certificate decides when that work peaks. AI agents now take part of the repetitive work at each stage, while offers stay with your staff.

This guide is for heads of admissions, access, recruitment and marketing in third-level education. It covers the definition, the Irish funnel and its metrics, a worked example you can redo with your own numbers, the strategies that move each stage, and a task-by-task view of what an AI agent can take off your team.

### Where the term comes from

The term is American. Its best-known definition comes from AACRAO, the North American association for registrars, admissions officers and enrolment managers. An [AACRAO Consulting workshop deck on strategic enrolment management](https://studentlife.media.uconn.edu/wp-content/uploads/sites/4161/2026/05/SEM-Plan-Core-Concepts.pdf), published by the University of Connecticut, quotes Bob Bontrager: SEM "is a concept and process that enables the fulfillment of institutional mission and students' educational goals." The deck stresses that it is not the work of a single unit and not solely about recruitment and marketing.

For an Irish team, that means the first conversation should capture whether the prospect is a school leaver, a mature applicant, a further education (PLC) student, an access applicant or an international student, plus the course and the year of entry, so each enquiry counts against the right target. Capturing those details in conversation and passing them to the CRM is the step Skolbot handles.

### What enrolment management covers

| Function | Question it answers | Typical owner |
|---|---|---|
| Marketing and recruitment | Who hears about us, and who enquires? | Recruitment and schools liaison |
| Admissions | Who applies, and who receives an offer? | Admissions, with CAO for most undergraduate courses |
| Access and supports | Who needs an access route or support to take up a place? | Access office, disability service |
| Acceptance and registration | Who accepts, registers and arrives? | Admissions and student services |
| Retention | Who continues into second year? | Student success and academic departments |

## The Irish funnel and its standard metrics

The Irish funnel runs from enquiry to CAO application, Change of Mind, offer round, acceptance and registration. Offers depend on CAO points calculated from Leaving Certificate results, and some courses add a portfolio, interview or test. Mature, international and access applicants follow other routes, so record the route on every enquiry.

| Metric | Formula | Notes |
|---|---|---|
| Enquiry to application | CAO applications listing your courses from the enquiry cohort ÷ enquiries | Needs enquiries recorded with a source and route |
| Offer rate | Offers ÷ applicants listing the course | Driven by points and places, not by service |
| Acceptance rate | Acceptances ÷ offers | The main conversion measure |
| Registration | Registered ÷ accepted | Captures students who accept but do not register |
| Melt | Accepted but not registered ÷ accepted | Measure your own, by route |

### Melt: the research behind the term

"Summer melt" comes from US research. Benjamin Castleman and Lindsay Page, in [their 2014 study in Social Science Quarterly](https://econpapers.repec.org/RePEc:bla:socsci:v:95:y:2014:i:1:p:202-220), found summer melt rates ranging from 8 to 40 percent among college-intending US high school graduates. Their [book with Harvard Education Press](https://education.brown.edu/publications/summer-melt-supporting-low-income-students-through-transition-college) states that anywhere from 10 to 40 percent of students presumed to be headed to college fail to matriculate at any institution that autumn. These are US figures; use them as a reason to measure your own loss between acceptance and registration. Our guide to [reducing melt after an offer](https://www.skolbot.ai/en-IE/blog/reduce-summer-melt-ai-agent-admitted-students) covers the tasks in that window.

## Worked example: how each stage rate changes registered students

Registered numbers are a product of rates, so a small change early in the funnel compounds:

**Registered = enquiries × enquiry-to-application rate × offer rate × acceptance rate × (1 − melt rate)**

The figures below are an illustration with stated assumptions, not data from any institution. Replace them with your own.

| Assumption | Value |
|---|---|
| Enquiries in the cycle | 10,000 |
| Enquiry to application | 20% |
| Offer rate | 60% |
| Acceptance rate | 33% |
| Melt (accepted, did not register) | 10% |

Step by step: 10,000 × 20% = 2,000 applications. 2,000 × 60% = 1,200 offers. 1,200 × 33% = 396 acceptances. 396 × 90% = **356 registered**.

| Change (one point) | Registered | Gain |
|---|---|---|
| Enquiry to application 20% → 21% | 374 | +18 |
| Acceptance 33% → 34% | 367 | +11 |
| Offer rate 60% → 61% | 362 | +6 |
| Melt 10% → 9% | 360 | +4 |

The gain follows the relative change. One point on a 20% rate is a 5% relative lift; one point on a 90% registration rate is about 1%. In a points-based system the offer rate is set by places and results, so treat it as outside your service work. To redo the calculation, take last cycle's counts by route, compute each rate and test a one-point change at each stage.

## Enrolment management strategies by stage

| Stage | Strategy | Measure |
|---|---|---|
| Planning | Set targets by route, course and year of entry | Targets met by segment |
| Enquiry | Answer the same day, and plan for peaks around results | Time to first answer; enquiry to application |
| Change of Mind | Explain courses and alternatives clearly before the window closes | Applications listing your courses |
| Offer rounds | Answer acceptance questions fast and route undecided students to an adviser | Acceptance rate |
| Fees and grants | Explain the Student Contribution and link to SUSI for grants | Acceptance among offer holders who ask about cost |
| Acceptance to registration | Keep contact on registration, supports and accommodation | Melt |

For conversion in more depth, see our guides to [yield management](https://www.skolbot.ai/en-IE/blog/yield-management-student-enrolment-schools) and to [a declining student pipeline](https://www.skolbot.ai/en-IE/blog/declining-student-pipeline-private-schools).

### An enrolment plan checklist

| Test | Pass | Fail |
|---|---|---|
| Targets | Numbers by route and course | One headline number |
| Definitions | Each stage defined in writing, shared by every office | Offices count applicants differently |
| Ownership | Each stage and handoff has a named owner | Nobody owns acceptance |
| Data | Enquiry, application, offer, acceptance and registration counts by cohort | Only offers and registrations |
| Links | Tied to the academic plan, budget and marketing plan | A standalone document |
| Review | Reviewed after each CAO cycle, changes recorded | Written once, then filed |

## Can AI automate repetitive tasks in the admissions office?

Yes, an AI agent can automate many repetitive admissions tasks: answering routine questions from approved content, collecting a few qualifying details, booking open day places and writing a summary into the CRM. It should not predict offers from points or decide on access supports.

| Admissions office task | Can an AI agent take it? | Condition | What stays with staff |
|---|---|---|---|
| Answering course, entry requirement and fee questions | Yes | Approved, current pages only, source shown | Owning the source content |
| Collecting route, course and year of entry | Yes | Asked in conversation, stored in agreed CRM fields | Interpreting the answers |
| Booking open days and adviser calls | Yes | Connected to your event calendar | Hosting the event |
| Creating the CRM record with a chat summary | Yes | Field mapping agreed and tested | Follow-up with context |
| Handling the results-week surge of routine questions | Yes | Content updated before results | Callers who need guidance |
| Explaining how CAO points work | Partly | Published information only, link to CAO, no predictions | Entry decisions |
| Offers and restricted-entry decisions | No | | Admissions staff |
| Access, disability and hardship cases | No, hand off at once | | Access office and trained staff |

See also [which admissions tasks to automate with AI](https://www.skolbot.ai/en-IE/blog/which-admissions-tasks-to-automate-with-ai) and our overview of [AI agents in higher education recruitment](https://www.skolbot.ai/en-IE/blog/ai-agents-student-recruitment-higher-education).

## Enrolment management software and where an agent fits

A student enrolment management system is usually a CRM at the centre, with CAO and direct application feeds, event tools and AI agents around it. The agent feeds the CRM. Salesforce and HubSpot are among the CRMs Skolbot connects to: the AI admissions agent on your website hands qualified enquiries into them, as described on our [integrations page](https://www.skolbot.ai/en-IE/integrations). Our [CRM comparison for higher education](https://www.skolbot.ai/en-IE/blog/crm-higher-education-comparison) covers how to choose a platform.

## What stays human

- Offers, and decisions on restricted-entry courses.
- Access and disability supports, including DARE and HEAR style schemes.
- Funding hardship and personal circumstances.
- Conversations with a student in distress, which need immediate handoff.
- Complaints and appeals.
- Relationships with guidance counsellors, schools and families.

## Risks to manage

**Wrong answers.** A wrong CAO date, a wrong points figure or an invented scholarship can harm a student who acts on it. Restrict the agent to approved content, show sources and appoint an owner.

**Data protection.** GDPR applies directly in Ireland, supported by the Data Protection Act 2018, and the [Data Protection Commission](https://www.dataprotection.ie/) is the supervisory authority. Have a lawful basis, be transparent, collect only what you need, set retention periods, run a data protection impact assessment and check whether sub-processors transfer data outside the EU. Many prospects are minors before the Leaving Cert, which calls for particular care. Tell people at the start that they are talking to an automated system.

**Accuracy of public information.** [Quality and Qualifications Ireland](https://www.qqi.ie/) (QQI) expects information for learners to be accurate, so the agent's descriptions of programmes and awards must match what is validated and published.

## A rollout approach that avoids the CAO crunch

1. **Choose one scope**, such as one school or questions before the CAO deadline.
2. **Fix content first**: align course pages, entry requirements and fee information.
3. **Agree handoff rules and CRM fields** with advisers.
4. **Test with staff**, including points, Change of Mind and mature-applicant questions.
5. **Launch on selected pages and review weekly.**
6. **Measure against your funnel**, compared with the same cohort last cycle.

Launch well before results week, not during it.

## Where Skolbot Chat fits

Skolbot Chat is a website agent that answers from the institution's own content, qualifies prospective students, books appointments and sends the record to the CRM with a summary for the adviser who follows up. WhatsApp is available as an optional module. Evaluate it as you would any supplier, against your own content and your own data protection review.

## FAQ

### Is enrollment management an Irish term?

Not typically. Irish institutions speak about admissions, access and recruitment. The American term is still useful because it treats the whole journey as one connected process with shared targets.

### What is strategic enrolment management?

It is enrolment management with a planning layer: targets tied to mission, academic plan and budget, reviewed each cycle. AACRAO uses Bob Bontrager's definition of it as "a concept and process that enables the fulfillment of institutional mission and students' educational goals."

### Can an AI agent tell applicants if they will get a place?

No. It can explain published entry requirements and how CAO points work, but it should not predict offers.

### Does GDPR apply to our chat agent?

Yes. It processes personal data, so you need a lawful basis, transparency and safeguards. Speak to your data protection officer before launch.

### Will an AI agent replace admissions staff?

No. It handles routine questions so staff can focus on advice and the decisions that need judgement.

<a href="/en-IE/demo" className="cta-button">Try Skolbot on your school in 30 seconds</a>
