The school admissions funnel: eight stages from inquiry to enrollment

A private school's enrollment season almost always begins as a list. Parents writing in on WhatsApp, mothers asking questions through Instagram DMs, people who saw the sign outside and called, friends referred by current parents. In most institutions these names collect in one place: a spreadsheet, a notebook, or the conversation list on the admissions officer's phone. As the list grows, one question becomes harder to answer: "How many families are seriously considering us right now, and how many of them are waiting on a step from us?" A list of names cannot answer that, because on a list everybody looks the same. But a family that has already walked the campus is not the same thing as a number that asked about fees three months ago and went quiet. This article describes how to turn prospective-parent follow-up into a school admissions funnel: the eight stages, the exit criterion between them, the standard lost-reason codes used at closing, and the funnel review you should run once a week.

Quick answer

Why a list of names has to become a funnel

The problem with a list of names is not missing information; it is missing order. If there are thirty families on the list, the admissions officer decides who to call from memory, and memory promotes whoever was spoken to most recently. The result is familiar: the family that toured the campus and is waiting for an offer waits three more days, while a brand-new number whose needs are still unclear gets called twice.

A funnel fixes this in three ways. First, the stage itself tells you what each family is waiting for; what a family in "Offer Presented" expects from you is not ambiguous. Second, once you set an exit criterion for each stage, you work with events that happened rather than vague impressions like "they seem interested". Third, once you code your losses, the end-of-season sentence stops being "enrollments were low" and becomes something you can act on, such as "twelve of thirty-four losses were about the bus route".

The funnel depends on the institution type. Institutions that select Private School (K12) or Nursery & Preschool during setup get the eight-stage funnel built around campus tours and written offers. Test-prep and study centers, language schools, and general training centers get their own seven-stage education funnel built around the trial lesson: New Interest, Informed, Demo / Trial Lesson, Awaiting Enrollment, Awaiting Payment, Enrolled, Dropped. That funnel has no separate tour stage; in those types the decision turns on the student sitting a trial lesson, not on the family walking the building.

The eight stages and their exit criteria

The table below is the backbone of the funnel. The right-hand column matters most: the event that must have happened for a prospect to move on.

A few stages need explaining in practice.

The difference between New Prospect and First Contact is the difference between one-way and two-way. A prospect does not count as advanced just because a parent wrote and you answered; if the parent has not written back, they are still a New Prospect. Preserving that distinction is what lets you know, mid-season, how many of your "fifty prospects" are actually in a conversation. In the private school and nursery setups, WhatsApp, Instagram DM, Messenger and Telegram replies land in an approval queue as drafts by default, so a staff member sees and sends the text of that first contact. Web chat and email sit outside that queue, so a first contact there goes straight from the AI. You can turn this off from Panel > AI > Advanced Settings. We cover the speed side of this stage separately in first response time at a school.

A tour booking is a line on the calendar, not an intention. "We'll drop by next week" is not enough to move to this stage. Once the appointment is created, reminders go out automatically one day before and two hours before; if the parent cannot make it, they reschedule or cancel themselves from the appointment link. Until both of those exist, the prospect should stay in First Contact, because a promise with no reminder does not convert at the same rate as an appointment written into the calendar.

Tour Completed is the real hinge of the enrollment season. After a family has seen the building, the classroom, and the teacher, the tone of their questions changes; the conversation moves from "which level are you looking at" to "how does the sibling discount work". The most common mistake at this stage is letting a family leave the tour without an offer. If there is no document in the family's hands after the tour, the only thing they can compare is their memory.

The difference between Offer Presented and Decision Stage is who holds the decision. Once the offer is presented, the ball is with the family. The moment they say "we'll talk it over and come back within two weeks" or "their father wants to see it too", a decision window has opened, and the follow-up rhythm is built around that window. The decision-timing field exists precisely for this: your reminder should go out before the window closes, because anything sent afterwards is late.

Closing works in two directions, and both of them are closings. Enrolled is marked as won, Lost is marked as lost. Moving a family into the Lost stage is not an admission of failure; it is the most valuable data source in your funnel — as long as it is closed with a code.

Stages also advance on their own

There is one fact you need while reading the table above: your team is not the only thing moving prospects between stages. After every incoming message the AI classifies the conversation and maps the sales stage it infers onto one of your funnel stages, moving the record there. This automatic advance runs on every message; it is not a setting you switch on.

The mapping does not look at stage NAMES; it spreads the conversation's progress fraction proportionally across the funnel's six non-terminal stages. In the school funnel that has a concrete and surprising consequence: when the AI works out that a tour appointment has been booked, it moves the record straight to "Tour Completed" — even though the tour has not happened yet. So when you look at the funnel you may find families in the "Tour Completed" box who have never walked the building. That is the exact opposite of the table's "the tour actually happened" criterion, and you should read the funnel knowing it.

That is why the table above is not an automation spec but your team's weekly review yardstick: the AI produces a first guess, and you compare it against reality and correct it. Writing the criteria down is what makes two different admissions officers see the same family at the same stage, and what makes the weekly stage distribution reflect reality.

One last warning: a stage correction you make by hand is not permanent. When the next message arrives the AI reclassifies and may move the record again on its own calculation. The one exception is closing: once a record is marked "Enrolled" or "Lost", the automatic advance leaves it alone. So mark the closing decisions you have made without delay.

The fields that should be filled on every prospect

The stages answer "where"; the fields answer "why". In a private school setup, the prospect record comes with the following fields:

The AI can extract all of these fields from the conversation. When a parent writes "my daughter moves up to 5th grade in September, does your bus come to Ataşehir", the level, the target grade, and the transport need all come out of the message itself. The practical benefit is that filling the record does not require a separate form-filling round.

Two fields are deliberately outside that behavior. The student's name and the student's year of birth are not extracted from the conversation automatically; only an authorized member of staff records them, with the parent's confirmation. In the nursery and preschool setup, the child's name, the child's year of birth, and the special-agenda field for the meeting work the same way. This is a deliberate boundary that keeps information about a child from entering the record through an inference. The AI does not ask for a child's identity or health information on its own in any case; if a parent opens a sensitive topic, it leaves the detail out of the record and defers it to the face-to-face meeting. We cover the data side in full in parent communication and data protection.

The source and referrer fields deserve separate treatment, because they are the two gaps most regretted at the end of a season. If source is empty, the question "did Instagram work this year" has no answer; you know the enrollment count but not where it came from. If referrer is empty, the school's strongest channel stays invisible: its current parents. In inTusell, a prospect can be linked to the person who referred them; once that link exists, the referral chain becomes visible in the panel, and when the prospect turns into an enrollment, a thank-you message goes to the referrer. Knowing who brought whom lets you build the end-of-season thank-you list from records rather than assumptions.

One caution about level: level is a field, not a targeting criterion. When you send a bulk update, targeting is done through the warmth segments (hot, warm, cold) or through the tags you apply to contacts. If you want to write only to high school prospects, you need to mark those parents with a level tag. Sends go out as a separate message to each recipient — parents never see each other's details — and a single run reaches at most 500 people.

Closing discipline: lost-reason codes

Writing free text while moving a prospect into the Lost stage makes that loss invisible in reports; notes like "didn't work out" or "never came back" cannot be counted. A standard code set is used instead:

There is a reason the location code stands on its own. One of the most frequently heard real reasons in a school or nursery funnel is "your bus doesn't come to our side" or "it's too far from home". That is not a product mismatch: the family liked the school, the distance stopped them. Writing both into the same code makes closing analysis misleading, because the actions they imply are opposites. A product mismatch means revisiting your messaging or your target audience; a location loss means revisiting the bus route. Because the code is separate, at the end of the season you can say something like "twelve families did not enroll because of transport, and all of them are from the same two neighborhoods" — and the answer to that sentence is a new route.

There are matching codes for the enrollments you win as well: the price fitting, the timing, the trust built, the value of a recommendation, the ease of the flow, and the strength of the relationship. Filling in won reasons never becomes as habitual as filling in lost reasons, but knowing what works is worth as much as knowing what does not.

The weekly funnel review

The funnel works when it is reviewed once a week, on the same day, in the same order. Three questions are asked in that review.

First question: how many prospects are in each stage? In a healthy funnel the count falls as the stages advance. If the count is piling up at one stage, that is where the blockage is. The two most common pile-ups are these: a build-up in First Contact means conversations are not being converted into tours; a build-up in Offer Presented means offers are being presented but not followed up.

Second question: how long have prospects been sitting in their current stage? Aging matters as much as the count. A family that has been in Decision Stage for three weeks is not really in Decision Stage; it has either been quietly lost or forgotten. Set an upper limit for each stage — five working days for Offer Presented, for example — and each week either advance the records that exceed it or close them with a code. The most harmful state a funnel can reach is looking full while being padded with dead records.

Third question: which codes did this week's losses fall into? A single week does not produce a meaningful distribution, but a pattern emerges after four. If response_delay and handoff_failure are stacking up, the problem is not in the funnel but in the speed of the operation. If price is stacking up, look at how the offer is being presented. If location is stacking up, look at the bus plan. If no_response is stacking up, look at the follow-up rhythm.

The most concrete benefit of asking these three questions regularly is that you can intervene while the season is still running, rather than after it ends. A blockage spotted in April can be fixed; a blockage spotted in August is only a note for next year.

What it isn't

• It is not a student information system. The funnel manages prospective-parent follow-up; it does not keep grades, report cards, attendance, class timetables, or official student records. It is not a replacement for e-Okul or similar official systems.
• It is not a payment system. The Enrolled stage records that the institution's own registration process is complete; inTusell is not a party to payments.
• A ready-made funnel is not a ready-made process. The eight stages are created in your account; defining which event earns which transition, reviewing the records, and running the weekly review are your team's job. The accuracy of the funnel depends on that discipline.
• It is not a tool for promising places or queue positions. On availability, the AI does not say things like "I've reserved your place" or "only a few places left"; if you are full, it says so honestly, explains the waiting list, and does not promise a date or a position.
• The institution type cannot be changed from the panel. It is chosen only during setup; what you change under Settings > Company Settings is the sector, and when the sector changes, the sub-type is cleared if it does not belong to the new sector.

Frequently asked questions

How many stages does the school admissions funnel have, and is it set up automatically?

Eight: Yeni Aday (New Prospect), İlk Temas (First Contact), Tur Randevusu (Tour Booked), Tur Yapıldı (Tour Completed), Teklif Sunuldu (Offer Presented), Karar Aşaması (Decision Stage), Kesin Kayıt (Enrolled — marked as won) and Kaybedildi (Lost — marked as lost). If you selected Private School (K12) or Nursery & Preschool as your institution type during setup, this funnel is created in your account automatically. Test-prep centers, language schools, and general training centers get a seven-stage trial-lesson funnel instead (New Interest, Informed, Demo / Trial Lesson, Awaiting Enrollment, Awaiting Payment, Enrolled, Dropped) — that funnel has no tour stage.

When should a prospect move to the next stage?

Every transition should have one observable criterion: the conversation became two-way for First Contact, there is a tour appointment with a date and time on the calendar for Tour Booked, the tour actually happened for Tour Completed, fees and terms were shared in writing for Offer Presented, and the family gave you a decision window for Decision Stage. The criterion is an event that happened, not an impression. But stages also advance on their own. After every message the AI classifies the conversation and moves the record to a stage, and because that mapping is proportional, a record can jump to "Tour Completed" as soon as a tour is BOOKED — before the tour happens. So the criteria are your team's weekly review yardstick: you compare the AI's guess against reality and correct it. A correction you make by hand can be recalculated on the next message; only the Enrolled and Lost closings are preserved.

Why should the source and referrer fields be filled in for every prospect?

The source field ships as a picklist in the school preset: website, instagram, whatsapp, referral, signage/neighborhood, fair/event, scholarship exam, other. Without these two fields you cannot say at the end of the season which channel produced enrollments or which family came through whose recommendation. The AI can extract these fields from the conversation; the referrer field is marked as requiring confirmation. When a referrer is linked to the record, the referral chain becomes visible in the panel, and when the prospect is won, a thank-you message goes to the person who referred them.

Does the AI ask for the student's name or year of birth on its own?

No. In the school preset, the student name and student year of birth fields are defined so that they are not extracted from the conversation automatically; only an authorized member of staff records them, with the parent's confirmation. In the nursery and preschool type, the child name, child year of birth, and special-agenda fields work the same way. The AI does not ask for a child's identity or health information on its own; if a parent shares something sensitive, it does not record the detail and leaves the topic to the face-to-face meeting.

Which codes are available when closing a prospect as lost, and why is location a separate code?

The standard lost codes are: price, timing, trust, competitor, product_fit, response_delay, handoff_failure, no_response, location, and other. Location is its own code because one of the most frequently heard real reasons in a school or nursery funnel is that the bus route is too far away. That is not a product mismatch: the family liked the school, the distance stopped them. Putting both into the same bucket makes closing analysis misleading.

Can I change the institution type later from the panel?

No. The institution type is chosen only during setup; there is no selector to change it afterwards in the panel. What you can change under Settings > Company Settings is the sector, and when the sector changes, the existing sub-type is cleared if it does not belong to the new sector. If you picked the wrong institution type, write to support.

Next step

Building the funnel is not a week's work; the real work is writing down the criteria and putting the weekly review on the calendar. Because the speed of first contact determines the top of this funnel, we suggest reading first response time at a school next. For the parent data, consent, and bulk-messaging side, see parent communication and data protection. To shape the AI with your school's knowledge base, tone, and boundaries, how to train the inTusell AI is your starting point.

You can review all the lines of the education solution on the education solution page, see package details on the pricing page, and reach the other articles through the blog.

If you would like us to set the admissions funnel up around your own season together, Get a demo or write to hello@intusell.com; we can settle the stages, the exit criteria, and your first weekly review in 30 minutes.