
TL;DR
- Enrollment automation is software that moves a student from admitted to cleared-to-start without staff pushing each step by hand. It runs the prerequisite, eligibility, document, placement and registration hold checks that usually sit in separate office queues, asks the student for whatever is missing on the channel they read, and sends judgment calls to the right reviewer.
- Summer melt research by Castleman and Page puts the share of college-intending students who never enroll at 10 to 40%.
- A good automated check compares what the student told you against what your records already hold, then asks the student only for what is missing.
- Exceptions such as course equivalencies and prerequisite overrides go to a named person with the full context attached.
- The SIS stays the system of record. The automation runs the conversation, the routing and the nudges around it.
A student gets admitted in April. She says yes in May.
By August she still has a missing transcript, an unread email about a placement test and a registration hold she does not know exists. In September her seat is empty.
Enrollment automation is the use of software to run the steps between an admission decision and a student's first class without a staff member pushing each one by hand. That covers prerequisite checks, eligibility rules, document collection, placement scores, registration holds and the reminders that tie them together. In EdTech and higher education, it is the part of enrollment where intent is already won and students still get lost.
This guide walks through what those checks cover, why manual queues leak students, how an automated prerequisite check should work step by step, and where people still need to make the call.
Where does enrollment stall after a student says yes?
Enrollment stalls in the gap between yes and day one. Admissions teams measure applications, offers and deposits, and speed to lead gets most of the attention.
The funnel looks finished at the deposit. The student's to-do list is just starting.
That list is long. It runs across several offices, and each office owns one piece of it.
- Academic readiness. Prerequisite courses, placement test scores and transfer credit.
- Paperwork. Transcripts, ID, residency proof and, for international students, visa documents.
- Account state. Registration holds for unpaid balances, missing immunization records or incomplete forms.
- Money. Aid forms and a payment plan, both of which can block registration.
Researchers call the loss in this stretch summer melt. In their 2014 book Summer Melt, Benjamin Castleman and Lindsay Page found that 10 to 40% of college-intending students fail to enroll in the fall after being admitted. A National Summer Learning Association brief on their work with uAspire students put melt at 22% among the lowest-income students, and close to 40% among students headed to community college.
The deposit is where admissions stops counting and where most of the enrollment work begins.
What do prerequisite and eligibility checks actually cover?
Prerequisite and eligibility checks confirm that a student is allowed to take the courses they plan to take, and is cleared to register at all. They answer two separate questions, and it helps to keep them apart.
Course prerequisites
A course prerequisite is a course, score or skill a student must have before enrolling in another course. Calculus needs algebra.
A nursing clinical needs anatomy. Some courses allow a corequisite instead, which means the student takes both in the same term.
The check sounds simple until the evidence lives in three places. The grade might be in your SIS, in a transcript PDF from another college, or only in what the student tells an advisor.
Eligibility to register
Eligibility checks ask whether the student can register for anything yet. A hold, a missing document or an unpaid balance can block a student who meets every academic prerequisite.
In the University of South Carolina's non-registered student initiative, 31% of eligible students who had not registered and carried a hold had two or more holds. That program tracked 6,454 students between December 2021 and August 2024. One cleared hold does not mean a cleared student.
Why do manual queues and fixed form logic leak students?
Manual queues leak students because each office works its own list on its own schedule, and the student sees none of it. A transcript waits for a records clerk.
A placement score waits for testing. The student waits for an email that sits unread in a school inbox they rarely open.
Transfer credit shows how slow this gets. One vendor's roundup of published turnaround times found the fastest schools evaluating credit in 24 to 72 hours and the slowest taking 6 to 8 weeks. Every week of that wait is a week a student cannot plan a schedule.
The GAO found that students who transferred between 2004 and 2009 lost about 43% of their credits on average.
Fixed form logic fails in a different way. A web form can require a field, but it cannot notice that a student uploaded the wrong year's transcript or ask why a placement score is missing.
It accepts the form, or it rejects it, and either way nobody follows up.
Students rarely quit enrollment on purpose. They stop when nobody tells them what is missing.
How does enrollment automation run a prerequisite check?
Enrollment automation runs a prerequisite check as a loop with five steps, closer to a dynamic process workflow than a static form. It reads what you already know, finds the gap, asks the student only about that gap, and keeps going until the student is cleared.
- Pull the requirements. For each course on the student's plan, read the prerequisites, corequisites and minimum scores from the catalog or SIS.
- Match against records. Compare those requirements with what the SIS already holds, such as grades, test scores, transfer credit and holds.
- Add what the student declared. Students often tell you things your records do not yet show. "I took AP Calculus." "My transcript is coming from my old college next week."
- Ask for only what is missing. Send one specific request, on the channel the student actually reads, for the single item that blocks them.
- Clear or route. When the evidence lands, mark the requirement met and move on. When it needs judgment, send it to the right person.
Step 3 is where most systems go blind. A declared fact from a WhatsApp chat or an advisor call usually lives in someone's notes, not in the record the check reads.
Zigment's Conversation Graph keeps one timeline per student across chat, WhatsApp, email and calls, so a declared fact like "my transcript is on the way" becomes a signal the check can act on.
| Requirement | Evidence on file | What the student is asked |
|---|---|---|
| Algebra before Calculus I | Transfer transcript not yet evaluated | Upload the transcript, or confirm the college name so the request can go there |
| Placement score of 70 | No score on file | Book a placement test slot from the next three dates |
| No registration holds | Immunization hold active | Upload the immunization record, with a link to the clinic |
| Payment plan in place | Aid form incomplete | Finish the two missing sections of the aid form |
What does AI automation for student enrollment add?
AI automation for student enrollment adds the parts a rules engine cannot do on its own. It reads the student's reply in plain language, works out which requirement it answers, and checks whether the attachment is the right document.
It also notices intent. A student who writes "not sure I can afford this anymore" needs a different next step than one who writes "sending it tonight."
The catalog says what is required.
The automation decides how to ask, when to ask again and when to stop asking and hand over. That is the difference between a fixed sequence and adaptive journey orchestration.
Ask for one missing item at a time, on the student's channel, and the checklist gets finished.
Which exceptions still need a person, and how do they get routed?
Some checks need a decision that only a faculty member or registrar can make. Enrollment automation should send these to the right person fast, with everything they need to decide in one place.
- Prerequisite overrides. A student without the formal prerequisite asks to enroll based on work experience or a portfolio.
- Course equivalencies. A transfer course does not match anything in your equivalency table yet.
- Conflicting evidence. The student says they passed a course, but the transcript shows a withdrawal.
- Policy edge cases. International credentials, late admission or a name mismatch across documents.
Each routed case should carry the requirement, the evidence on file, what the student said and a suggested decision. The reviewer reads one case and decides. The decision then flows back into the loop, and the student hears about it on the same channel they used to ask.
This is how escalation should work in practice. The automation handles the routine checks on its own, and the configuration decides which case types go to which reviewer.
How should enrollment automation nudge students until they are cleared?
Enrollment automation should nudge on the student's channel, about one specific item, at a pace tied to the start date. The research on melt points the same way. In the uAspire work summarised by NSLA, outreach on placement tests, course selection and aid forms reduced summer melt by 20%, and that text messages were among the most effective ways to reach students.
Direct outreach to students who are eligible but not registered works too. At the University of South Carolina, advisor outreach between 2021 and 2024 cut the share of eligible students who had not registered from 8.71% to 3.53% per semester.
A workable cadence looks like this.
- Day 0. Send a single message listing the one item that blocks registration and how to fix it.
- Day 3. If nothing has arrived, send a shorter reminder with a direct link or a reply option.
- Day 7. Switch channel, from email to WhatsApp or SMS, or the other way around, keeping one thread across channels.
- 14 days before start. Hand the case to an advisor with the full history attached.
Every nudge should stop the moment the item lands. For the science behind timing and tone, see our piece on automated nudging and student retention.
A nudge that names the exact missing item gets an answer. A generic reminder gets ignored.
Where does the SIS end and the orchestration layer begin?
The SIS stays the system of record. It holds the catalog, the grades, the holds and the final registration. Enrollment automation sits on top of it and runs the work around those records, meaning the conversations, the checks, the routing and the follow-up.
Most admissions automation stops at the deposit. This enrollment workflow picks up from there and runs until the student is in class.
You keep your SIS as it is, and every decision still lands in the record your registrar trusts.
Zigment is a Conversational Revenue Orchestration Platform for education and EdTech growth teams. It runs the student conversations on WhatsApp, web chat and email, keeps them in one timeline, and orchestrates the next step from what each student says and does. One EdTech platform running its enrollment conversations on Zigment handles more than 100,000 conversations a month and saw a 3x lift in conversion, with WhatsApp and web chat sessions that carry over from one visit to the next.
If you are mapping the full path from first enquiry to graduation, our guide to orchestrating the full student lifecycle covers the stages before and after this one.
What to automate first
Start with the check that blocks the most students. For most institutions that is one of three things.
- Missing transcripts, because they block both prerequisite checks and transfer credit.
- Placement tests, because they have a booking step students put off.
- Registration holds, because students often do not know they have one.
Pick one, run it for a single intake and measure how many students reach cleared-to-start before the first day, compared with the intake before.
Fix the check that blocks the most students first, then add the next one.
The students you already won
Most strategies to increase student enrollment aim at the top of the funnel. The students lost after they say yes are the cheapest ones to keep, because they already chose you. Mostly, they need to know what is missing.
So here is a question worth putting to your team this week. How many of last fall's admitted students never made it to a first class, and how many of them were waiting on something you could have asked for in one message?
If you want to see how one student timeline turns each of those checks into a clear next step, take a look at how the Conversation Graph works.