A private school can raise enrolment by following up every enquiry within one working day, with a reply that answers the family's real question and a clear next step such as a tour. AI drafts the replies, invitations and reports. People make every admissions decision, check every message and keep the funnel counts that show where families drop away.
Speed matters because families rarely enquire at one school. A parent who writes on a Sunday evening and hears from one admissions office on Monday morning and another on Thursday has already formed a view before the first tour. That is a process problem more than a marketing one, and it is one an admissions team of two or three can fix without adding staff.
Draw the admissions calendar before you plan any message
Admissions work comes in waves. Enquiries peak at certain points in the year, tours cluster before application deadlines, and offers and deposits follow a fixed order. A calendar that shows the cycle tells you when speed matters most and when to prepare templates. Make one page with a row for each month and four columns: what families are doing, what the school sends, who is responsible, and what is counted.
Here is an illustrative cycle for a school that takes applications for entry the following September. Replace the months with your own.
| Period | What families do | What the school sends |
|---|---|---|
| Autumn term | Enquire, book tours, attend open mornings | Prospectus link, tour invitation, open-day reminder |
| Late autumn to January | Submit applications | Acknowledgement, assessment-day details, deadline reminder |
| February to March | Attend assessments or interviews | Invitation, preparation note, thank-you |
| March to April | Receive decisions, accept or decline | Offer letters, deposit information, waiting-list notes |
| May to August | Settle, buy uniform, meet the class | Welcome pack, transition days, forms |
| Rolling | In-year enquiries | Same replies, faster, with current availability |
Pin the calendar where the team can see it, and add the dates when the office is busiest, such as the day after an open morning. Each peak is a test of the reply process. If you receive 40 enquiries the morning after an open day and the team can only answer ten that day, the rest will wait, and a message sent too late is often a lost family. Plan extra cover for those mornings, and have templates ready.

The Complete Excel Course with AI Integration
From formulas and pivot tables to Power Query, DAX, VBA and AI with Copilot, ChatGPT and Claude — hands-on from the first lesson.
By Sagnik Bhattacharya and Virginia Thorn, Primetime Emmy Award winner
Buy nowTime your replies for a month, then set a standard
You cannot improve what you have not measured. For four weeks, record for each enquiry the time it arrived and the time of the first human reply. Do not guess from memory; use the time stamps in your email or customer system. At the end, sort the times and look at three numbers: the middle value, the slowest 20%, and the number that took more than two working days.
Say the illustrative result is that 60 enquiries have a median first reply of 26 hours, the slowest fifth took more than three days, and eight enquiries were never answered because they landed in a shared inbox that nobody owned. Those eight are the quickest win: assign an owner and a daily check. Then set a standard you can keep, such as an acknowledgement within one working day and a full reply within two. Write it down and share it. A standard that is too ambitious will be broken in the first busy week and will teach the team to ignore it.
Let AI sort and summarise, not answer
Once enquiries come through, a chatbot can speed up the first step: reading the message, picking out the question and suggesting which template fits. Paste the enquiry with the child's name and any personal details removed.
Read this enquiry and return:
1. The parent's main question in one line
2. Any second question
3. Which of our templates fits: general information, fees, tour request, scholarship or bursary, in-year place, other
4. A line saying whether a staff member must check anything first (fees, availability, a child's needs)
Enquiry: "[paste with names removed]"
Do not draft a reply yet.
This gives the admissions officer a head start without letting the model speak for the school. A typical output flags that "bursary" in the message needs the finance office. It will also flag any mention of a child's special educational needs, a medical condition or a family circumstance, which should go straight to a named person rather than being answered from a template. If the output misses such a flag, tell the model to look harder for it and keep the human in charge.
Draft replies that answer the question and offer one next step
A good first reply does four things in under 150 words: thanks the parent, answers the question they asked, gives one useful piece of information they did not ask for, and offers a single next step. The next step should be a tour or a call, never a vague "get in touch if you have questions".
Keep a template library of eight to ten replies, one for each common enquiry, and let the chatbot draft a variant for each message. Here is a worked example. A parent writes: "Hello, we are thinking about a place for our daughter when she starts secondary school next year. Could you tell me about fees and whether you offer bursaries?"
Draft a reply (under 150 words) from the admissions office.
Facts to use:
- Current fees are listed on our fees page [link]; we will send the full schedule after a tour
- Bursaries exist; the application is separate and confidential, and the bursar will explain
- Tours run on Tuesday and Thursday mornings; the next two are [date] and [date]
Do not quote a figure that is not in these facts. Do not promise an outcome. Do not ask about the family's finances. Offer the tour as the next step. Use [first name] and [your name].
A useful draft is brief and courteous. Fix three things before sending: replace the placeholder links and dates, confirm that the bursary wording matches your published policy, and delete any sentence that implies a place or an award. A reply that says "bursaries are available for deserving families" overpromises and subtly judges the family; "bursaries are available, and the bursar can explain how the process works" is plain and fair.
What not to automate
Some messages must never leave without a person. These include anything about a child's needs, behaviour history or health; any mention of safeguarding; a complaint; a request about fees that departs from the published schedule; and every offer, rejection or waiting-list decision. Draft templates for these with a manager's approval and keep them out of any automatic reply. An automatic acknowledgement ("we have received your message and will reply by tomorrow") is fine, because it makes no decision and promises only what you can deliver.
Invite families to open days in a way they act on
An invitation has to answer five things quickly: when, where, how long, who they will meet and what to do. Add a line about parking or access if it is awkward. Send it twice, a fortnight ahead and two days before, and add a short message the morning after to those who came and to those who did not.
Write three versions of an open-morning invitation email (under 110 words) for a private school.
Facts: date [date], 9.30am to 12pm, tours of 40 minutes with the head of admissions or a senior pupil guide, a short talk by the head at 10.30, refreshments, no booking needed but booking helps us plan.
Versions: warm and short; a little more formal; a version for families who have already toured.
Do not claim any exam results, rankings or inspection judgements.
Use [first name] and [your name].
Pick the version that sounds like your school, and delete any figure or ranking you cannot source and date. The "morning after" message is the one schools forget. For those who attended, it thanks them, answers a question they raised and offers a one-to-one conversation. For those who booked and did not come, it says they were missed and offers the next date. Keep a simple list of both groups, because they are your warmest leads.
Counts that show where families drop out
Six numbers are enough: enquiries, tours or open-day attendees, applications, offers, acceptances, and the time of the first reply. For each cohort, work out the ratio at each step. Here is an illustrative cycle for a school with 60 places in an entry year.
| Step | Families | Ratio to previous step |
|---|---|---|
| Enquiries | 210 | |
| Tours or open-day attendance | 112 | 53% |
| Applications | 84 | 75% |
| Offers made | 72 | 86% |
| Places accepted | 54 | 75% |
To fill 60 places at those ratios, the school needs about 233 enquiries: 54 acceptances from 210 enquiries is one in about 3.9, and 60 x 210 / 54 is roughly 233. The most promising lever depends on where the steps are weak. A tour rate of 53% suggests reply speed and invitation clarity are worth fixing. A 75% acceptance rate may reflect families choosing another school after the offer, so a prompt, personal follow-up to each offer matters. Ask the chatbot to compute these ratios from the counts you paste in, and to show its arithmetic. Never include names, and never paste a spreadsheet with child-level data.
Compare each cycle with the same point in the previous one rather than with last month, because admissions is seasonal. If you recorded reply times, split the enquiries into "replied within one working day" and "later" and compare the tour rates. If the first group tours at, say, 62% and the second at 41%, you have a local reason to protect the standard. Those figures are illustrative; your own counts will differ, and a small sample can mislead, so look for patterns across several cycles.
Follow up every offer and every waiting-list family
The gap between an offer and an acceptance is where a school can lose a family it has already won. A parent who holds three offers will often decide on the strength of the last personal contact. A short phone call or a handwritten line from the head of admissions often does more than a second email. AI can prepare the call: paste the offer date, the deadline for acceptance and a line about what the family liked on the tour (with no names), and ask for a one-page call sheet with three questions and a closing line. The person then makes the call and updates the record.
Waiting-list families need care for a different reason. They are waiting for a decision that depends on other people, and silence feels like rejection. Draft a monthly update that says where the list stands in general terms, repeats how places are offered under your published criteria, and invites the family to say if their plans change. Never imply a rank or a likelihood you cannot back up, and never let the model estimate a family's chances. If a place opens, a person phones first and sends the written offer afterwards.
An offer follow-up email, drafted and checked
Draft a warm email (under 120 words) to a family who holds an offer for the entry year which they must accept by [date].
Facts: they attended a tour and an assessment morning; they asked about the music programme and the bus route; the deposit information is attached.
Include: congratulations, a reply on both questions using only "music lessons run before school and at lunch" and "the bus route list is attached", an offer of a call with the head of admissions, and the acceptance date.
No pressure language, no "limited places", no mention of other applicants. Use [first name] and [your name].
A usable draft is friendly and factual. Before sending, check that the acceptance date is correct and that the attached documents are the current ones. Replace any "we would love to have you" with the school's usual tone. If the email says anything about the child's ability or place in a ranking, delete it. A call from a person the same week completes the follow-up, and the outcome goes in the record.
In-year enquiries need their own speed
Families who move or change schools part-way through the year are often in a hurry, and they are less likely to wait for the next tour. In-year enquiries are small in number and high in value, because a ready-to-join family can fill a single empty seat. Keep an up-to-date list of places by year group, and reply with the current position, not last term's. Offer a short tour within the week, and tell the family clearly which documents you will need. A single "in-year" template, checked by the head of admissions, plus a daily glance at that inbox, covers most of the work.
For an illustrative sum: a school with six free places across years in the autumn term and two in-year enquiries a week could fill most of them within a term if it replies the same day and offers a tour within five days. A slower school, replying in three days, might lose half to schools that answered sooner. Compare your own numbers and see whether the in-year inbox deserves an owner of its own.
Parent testimonials, with consent
Parents' words carry weight in a prospectus or on a website, and they raise legal and ethical questions. Ask for written consent that states where the quote will appear, in which formats and for how long, and give an easy way to withdraw. Show the parent the exact wording before it is used, and never change it beyond a light spelling correction they have approved. Do not attach a child's name, photograph or year group without separate consent from the people entitled to give it. AI can help draft the request and tidy a form, but it must never write a testimonial, "improve" a real one or invent one. A made-up quote is a false statement, however plausible.
A short consent form can be drafted with AI and then checked by whoever advises you on data protection. It should say what is collected, why, who will see it, how long it is kept and how to withdraw. Review the form each year and delete quotes when consent lapses. Where families contact you from the EU, check the data-protection law that applies to you before you collect anything.
No automated decisions, in writing
Write a short rule and put it in the admissions handbook: AI may help draft messages, summarise information for staff and prepare reports. AI may not score, rank, shortlist or reject an applicant, and no decision is made on the basis of an AI output alone. Name the roles who decide, and keep a written record of each decision and its reason. This protects families, and it protects the school if a decision is queried. The tutorial on setting up human review for AI work shows a light way to do it without slowing the team.
Check your data rules too. Enquiries and applications contain personal data about children and families. Use a business-grade tool with data controls, limit what staff paste in, and keep prompts free of names and sensitive details. Ask your data-protection lead to approve the tools and the rule on pasting before the first message is drafted. The broader tutorial on AI for private schools: admissions, fees and parent communication covers fee reminders and parent messages in more depth, so this tutorial stays with enquiries and conversion.
A realistic first 30 days
Week one: assign an owner to every inbox and form, and start logging arrival and first-reply times. Week two: write the eight to ten templates and have the head of admissions approve them. Week three: start using AI to sort enquiries and draft replies, with a person sending every message. Week four: review the counts and the reply times, and decide whether the standard is realistic. Keep the weekly check to 20 minutes.
Along the way, expect some of the following problems:
- Templates that sound alike. If every reply opens "Thank you for your interest in our wonderful school", parents notice. Vary the first sentence with a detail from the enquiry.
- A confident wrong answer. A draft that says "we have places in the year above" when you do not is worse than a slow reply. Keep an availability sheet and update it daily in busy periods.
- Staff pasting too much. A busy officer pastes a whole email thread with a child's needs into a public chatbot. Training and a one-line rule fix this, and so does a business plan with data controls.
- Counting the wrong thing. Enquiries that are really spam or duplicates inflate the top of the funnel. Clean them out monthly.
To check it worked, compare the four weeks before the change with the four weeks after on two numbers: the median time to first reply and the share of enquiries that book a tour. An illustrative result: the median falls from 26 hours to 7, and tours booked rise from 27 of 60 enquiries (45%) to 33 of 62 (53%). That is encouraging, not proof, because a month in autumn and a month in spring draw different families. Repeat the comparison at the same point next cycle before you credit the change.
When that works, a family who writes on Sunday evening gets a clear, personal reply by Monday lunchtime, with two tour dates, from a person who is checking the facts. That will not suit every family, and it will not fill every place, but it removes the avoidable losses. If you would like help mapping this for your own admissions office, that is what my AI implementation consultation covers, and the tutorial on parents' evening booking systems shows the same approach for slots and reminders once families have joined.
Questions admissions teams ask about faster replies
How quickly should a school reply to an enquiry?
Set a standard you can keep, such as an acknowledgement within one working day and a full reply within two. Families often enquire at several schools, so the first clear answer has an edge. Measure your actual reply times for a month before you set the target, then publish it internally.
Can AI decide who is offered a place?
No. Admissions decisions should stay with named people, using your published criteria. AI can draft replies, summarise notes and format reports, but it should never score, rank or reject a child or family. Keep a human decision-maker and a written record of every outcome.
Do we need consent to quote parents?
Yes. Get written consent that says exactly where the words will appear and for how long, allow withdrawal at any time, and never edit a quote beyond what the parent agreed. Do not use pupil names or images without separate consent from the people entitled to give it.
What should the admissions team count?
Enquiries, tours or open-day attendees, applications, offers and acceptances, plus the time of the first reply. Those six figures show where families drop out. Review them monthly, and compare like with like across the same point in each admissions cycle.
Further reads
- How to Reply to Every Enquiry in Under Five Minutes With AI — The general habit of fast, checked replies.
- AI for Coworking Spaces: Enquiries, Tours and Room Bookings — How another tour-based business handles enquiries.
- AI Automation for Education: Five Workflows, Tools and Costs — Five education workflows to automate first, with costs.
- AI for a Music School: Lesson Notes and Parent Emails Worked Through — Parent emails for a school that teaches children.
- Is It Safe to Use AI With Children's Data in a Nursery? — Children's data rules that apply to any school team.
- Should You Let AI Auto-Post Replies to Google Reviews? — Why review replies need a person before they go live.
- Language School Marketing With AI: Agent Emails and Translations — A small language school's plan for niche positioning, agent updates, checked translated landing pages and social posts, with AI doing the drafting.
- FERPA and ChatGPT: Does the Law Allow US School Staff to Use It? — What FERPA allows when staff use ChatGPT, what never to paste, and a one-page staff rule with a redaction routine for a private school.
- How to Start a Language School: Accreditation, Agents and AI — What a new language school costs in year one, how accreditation and agents fit together, and where AI saves admin time from the first term.
- Starting a Microschool: Enquiries, Enrolment and Parent Updates With AI — The admin a new microschool needs for enquiries, enrolment and parent communication, with prompts, a worked week and the legal checks AI cannot do for you.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: the school's own admissions records and published criteria, data-protection guidance for schools (reader to check current requirements), and the help pages of the AI tool used.