A small swim school should buy swim school management software for level tracking, waitlists, make-up lessons and billing first, and treat AI features as an optional extra. Compare vendors on those four jobs using your own class data. AI that drafts messages or reports can help, but it should never replace the rules that place children.
Many vendor pages lead with AI because it sells. Yours will be judged by whether a parent can reschedule a missed lesson at 9pm without ringing the desk, and whether a child who passes a level lands in the right class. Check those before you look at anything labelled "smart".
The four jobs software must do before AI matters
Swim schools have an unusual shape: small groups, strict ratios, progression through levels, a term or rolling calendar, and families who miss lessons for illness and holidays. Software that suits a general fitness studio often misses that shape. Rank the following by how much staff time each one eats today.
Level tracking and class placement
Each child has a level, skills passed and skills still working on. The system should show a class list by level and coach, record a pass, and offer the child's next class. Jackrabbit's swim page, for one, lists "Student Skills Tracking" as a feature but describes it only with a screenshot, so ask a vendor to show you the actual screen: how do you define levels, can one child hold different levels for different strokes, and what happens when a child passes mid-term?
Waitlists
The waitlist decides how full your classes stay. Look for: waitlist by class and level, automatic offers when a space opens, a time limit to accept, and a record of who declined. A waitlist that only collects names is a list, not a feature. Jackrabbit's swim page does not mention waitlists at all, which is not proof the feature is missing; ask every vendor, that one included, to show the waitlist working on a live class.
Make-up lessons and credits
Illness is the biggest source of rescheduling. Vendors know this: Jackrabbit's swim page, for example, lists an "Absences & Makeups" feature aimed at cutting calls and emails about missed classes. The software should let parents report an absence, see eligible make-up slots (same level, space available, inside your window) and book one; staff should be able to override. This is where schools discover that a system books make-ups without checking ratios. Agree your policy first; swim school cancellation and make-up lesson wording helps you write it before you configure anything.
Billing
Billing in swim schools is recurring, by term or by month, often with sibling discounts, registration fees and pro-rated starts. Jackrabbit's swim page lists "Automated Billing" among its features, and most rivals offer something similar, so the label tells you little. Ask about failed-payment handling, how a make-up credit or a refund is applied, and what the parent sees on their statement. One more check: which fees come on top of the subscription, such as card processing or text messages.
Demo questions for a swim school, with what a pass looks like
| Need | Question to ask the vendor | Pass looks like |
|---|---|---|
| Levels | Show me a child passing Level 3 mid-term. | Next class offered in two clicks; history kept. |
| Waitlist | What happens when a place opens? | Automatic offer, deadline, next child if declined. |
| Make-ups | Can a parent self-book a make-up? | Only eligible slots shown; ratios respected. |
| Billing | How are sibling discounts and pro-rata starts handled? | Set once, applied automatically. |
| Parent view | What does a parent see on a phone? | Timetable, credits, payments, messages. |
| Data | Can I export all students and history? | A plain file, no extra fee. |
| Cost | Total monthly cost at 180 swimmers? | One written figure, fees listed. |
Weight the first four double. A system that scores badly on any of them will cost you hours every week, whatever AI it adds.
A worked example: a 180-swimmer school
An illustrative school runs six lanes, 22 classes a week and 180 children, with two staff on the desk part-time. Over a term, the desk handles roughly 90 absence reports, 40 make-up requests, 25 level-change requests, 30 waitlist enquiries and a stream of billing questions. Assume five minutes each across about 185 requests, so roughly 15 hours a term, plus a monthly billing run of two hours.
Suppose the chosen software lets parents report absences and book make-ups themselves, cutting 90 percent of those requests, handles waitlist offers automatically and reduces billing questions by a third. That removes around 11 hours of desk time a term, which is real but not dramatic. If the desk costs $18 an hour, that is about $200 a term from the core system alone. What remains is roughly four hours a term, worth about $70, so an AI add-on costing $50 a month ($150 a term) could not pay for itself on desk time even if it removed every remaining request. The working rule is to buy the core system because it saves staff time, and to ask any AI feature to beat its own fee on its own evidence.
Check the sum with your own counts. If your desk logs 300 requests a term and most are make-ups, self-service make-ups are the feature to prize. If most are billing queries, look hardest at statements and failed-payment handling.
Waitlist rules written out, with a filled-in example
Software will only run the waitlist you describe. Write the rules in plain sentences before the demo, so you can check each vendor can follow them. An illustrative set for a school with eight-child classes:
- A place that opens is offered to the first waiting child at the same level and an age within the class band.
- The offer is sent by text and email and lasts 24 hours. If it is not accepted, it moves to the next child.
- A child who declines twice drops to the bottom of the list.
- Siblings are offered places in the same time slot where possible, otherwise adjacent slots.
- Staff can bump a child up with a reason recorded.
Now read a vendor demo against that list. The common gap is the second rule: many systems notify everyone on the waitlist at once, and the fastest parent wins. That seems fair but breaks your queue and annoys the families who waited longest. Others do not support sibling logic at all. Each gap costs desk time, so record it on the scorecard.
Staffing, ratios and the lane plan
Class capacity in a swim school is set by the water, the coach and the child's age, not just a number in a box. Check whether the software can hold different maximum sizes for different levels (for example four in a beginner class, eight in a stroke-development class), tie a class to a coach, and flag a coach who is double-booked. Rota features are usually weaker than booking features. If your coaches change weekly, ask whether the system sends them their schedule and whether changes notify parents. Where coach management is thin, many schools run rotas in a separate shared calendar and accept the double entry. Be honest about that on your scorecard, because it is time cost the vendor will not mention.
Importing your students: a ten-minute test
Move 20 real students into a trial account and run through this checklist. Illustrative results are in brackets.
- Do names, dates of birth and parent contacts import without errors? (18 of 20 imported; two had merged surnames.)
- Do levels import as tracked fields or as notes? (Notes. A manual re-entry would take about 90 minutes at full size.)
- Can siblings be linked to one family account? (Yes, but only after import.)
- Does a pass update the class list? (Yes, after a refresh.)
- Can you export everything again as a file? (Yes, one file per table.)
An import problem found in the trial costs ten minutes. Found in January, it costs a weekend. If the vendor offers a free import service, ask to see it done on your sample, not described.
Comparing quotes fairly
Pricing is where small schools get caught, because headline figures omit extras. Ask each vendor for a written quote at your real enrolment, and put the lines side by side. An illustrative comparison at 180 swimmers, using made-up figures, not any vendor's prices:
| Cost line | Vendor A | Vendor B |
|---|---|---|
| Base subscription | $120 a month | $85 a month |
| Per-student fee | Included | $0.50 per student |
| Card processing | Standard rate | Standard rate plus $0.10 a payment |
| Text messages | 500 a month included | Pay per message |
| AI add-on | $40 a month | Not offered |
| Set-up and import | $0 | $150 |
| Estimated monthly total | $160 | About $200 |
The vendor that looked cheaper at $85 is the dearer once students, payments and messages are counted. The headline number is the least useful figure on the page. Also ask what the price becomes after the first year and whether the contract auto-renews.
A term notice, drafted with an assistant
You do not need AI built into the software to draft parent communication. A general assistant such as ChatGPT or Claude can do it, and you paste the result into the system. Example prompt:
Write a 120-word term-start message for parents of a swim school.
Include: term starts Monday 4 January; lessons weekly at the same
time as last term; absences must be reported by 10am on the day
through the parent portal; one make-up lesson per child per term;
fees are taken on the 1st. Tone: friendly and direct. Do not add
any policy not listed here.A realistic output reads well but may add a line such as "no refunds will be given", which you did not supply. The "do not add any policy" line reduces that, and a read-through catches the rest. This is why built-in message drafting is a modest selling point: it saves a minute, and it brings the risk of something going out unread.
Which AI features are worth testing
Jackrabbit's swim page mentions no AI features as of October 2026, and other vendors vary from month to month, so use these as test categories rather than a list of what exists.
- Message drafting: help writing term notices, waitlist offers and reminders. Useful, low risk, and you can get the same from a general assistant.
- Progress report drafting: turns checklist ticks into parent-friendly text. Useful if it works only from what the coach recorded; see the level-by-level progress report template for how to structure that.
- Chat or phone assistant: answers timetable and price questions. Worth testing if the desk is swamped at peak enrolment, risky if it can only quote and not book.
- Class-placement suggestions: proposes classes from level and age. Treat as a suggestion a coach approves, never as automatic.
- Forecasting and reports: predicts drop-off or demand. Needs a few terms of data before it says anything useful.
For each, ask: does it use my data, can I see what it wrote before it goes out, and what does it cost on top of the base price? If a vendor cannot answer in a demo, treat the feature as unfinished.
Testing an AI report draft: a realistic sample
Suppose the coach records for a seven-year-old: "Level 3. Floats on back 20 seconds. Kicks on front 10 metres with board. Breathing to the side not yet consistent. Needs to work on streamlined push-off." A good draft reads:
[first name] has made steady progress in Level 3. They float on
their back for 20 seconds and kick 10 metres on their front with
a board. Side breathing is not yet consistent, so we will keep
practising it. A streamlined push-off from the wall is the next
skill to work on.A bad draft adds "and will be ready for Level 4 next term" or "shows excellent water confidence". Neither is in the notes. That is the fix to apply: delete every sentence that cannot be traced to a coach's entry. A school that approves drafts without reading them will eventually promise a promotion that a coach never gave.
Before and after: the make-up lesson flow
Before: a parent texts the desk phone at 6pm, the message is seen the next morning, a staff member checks three class lists, finds an open place, replies at 10am, and the parent is out by then. After: the parent opens the portal, taps "report absence", sees two open Level 3 slots this week, books one, and the credit is recorded. The AI question is irrelevant; the workflow is. A school that automates this saves hours every term, and that comes from rules and a parent screen.
Parent communication: what to switch on first
Before any AI, switch on the plain automation most systems already have. Lesson reminders the evening before, an absence confirmation that shows the make-up options, a waitlist offer, a failed-payment notice with a retry link and a term-end message with next-term enrolment. These five messages account for most of what a desk sends. Set each to send from a named person's address so replies reach someone, and look at the open and reply patterns after a month: if parents reply "which pool?" to every reminder, add the pool and lane to the template. This kind of small fix often saves more time than any AI feature, and it costs nothing.
Where AI features go wrong in a swim school
- Invented policy: a chat assistant says a missed lesson can always be made up, when your policy allows one per term.
- Wrong level advice: a placement suggestion based on age alone puts a nervous six-year-old in a faster class.
- Safety wording: any assistant answering a question about a child's medical condition, a seizure or a water-confidence problem should hand over to staff, immediately.
- Privacy: children's names, ages and notes are sensitive. Ask where data is stored, whether it trains any models, and how it is deleted.
Put a sentence in your instructions that medical, safeguarding and injury questions go to a named person, and test that it does.
Small school, large school, different answers
Under 80 swimmers
A basic system with level tracking and online payments may be all you need. AI is a nice-to-have; a general assistant for drafting emails may cover it without a new subscription.
80 to 250 swimmers
Make-up self-service and waitlist automation are worth the most. Compare two or three vendors on those. AI chat is worth a trial only if enrolment weeks swamp the desk.
Multiple pools or franchise-style
Reporting across sites, role-based access and coach scheduling matter. AI is mostly reporting and forecasting; ask for a demo with your multi-site structure, not a single-site example.
Holiday blocks and intensive courses
Many schools run one-week holiday intensives alongside the weekly programme. Check that the software handles a short course with its own price, its own waitlist and a different capacity, without disturbing the weekly classes. Test one case: a child enrolled in a weekly Level 2 class who also books a holiday course. Do the two show on one timetable, and does billing treat them as separate charges? A system built for rolling weekly classes often struggles here, and you will be doing intensives by spreadsheet again.
A mistake that shows up in the second term
A school picks software on price, imports its students in September and discovers in January that levels were imported as text notes, not tracked fields. Every class list needs rebuilding by hand. The check is simple: before signing, import 20 real students into a trial, advance three through a level, and confirm the class lists update. Use what an API is and why it matters if you plan to connect the system to a website or accounting package, because import and export options decide how trapped you will be.
How to decide, in order
- List your last term's desk requests by type and count them.
- Score two or three vendors on levels, waitlist, make-ups, billing, using real student data.
- Get written total monthly cost at your enrolment.
- Only then ask what AI is included, and test each feature against your own desk counts.
- Run a one-term trial with a clear review date.
If after that you still want a second opinion, my AI implementation consultation covers this kind of choice; start from the contact page. For another small-business comparison with a similar structure, see pet grooming software with AI.
Swim school software questions
What is the most important feature in swim school software?
Level tracking tied to class placement. If a child passes a level, the system should show which classes they can move into and whether space exists. Without that link, front-desk staff track levels in a spreadsheet and the software becomes a payment taker. Test it with real children's records, not the sample data.
Does a swim school need AI to run waitlists?
No. A good waitlist is rules, not AI: first in line, matched to level and age, notified by text, with a time limit to accept. AI may help word the messages, but the matching logic should be explicit. Ask to see the waitlist rules screen in a demo.
Can AI write swim progress reports?
It can draft them from your level checklist and coach notes, which a coach then checks. Never let it add skills a child did not show. Keep the checklist as the source and treat the draft as a first pass.
How much should a small swim school spend on software?
Price varies with the number of students and add-ons, and many vendors list no prices openly. Compare total monthly cost including payment processing, text messages and any per-student fee, at your real enrolment, using a written quote rather than a headline figure.
Further reads
- Can AI Run Your Studio's Class Waitlist Automatically? — How AI can help fill waitlisted class places.
- Dance Studio Registration: Enrolment Emails and Waitlists With AI — Enrolment and waitlists in another children's class business.
- Martial Arts School Software: Gradings and Parent Updates Compared — Gradings work like swim levels; compare the software.
- Riding School Software: Lesson Booking, Rider Notes and AI Reminders — Compare riding school software on horse-and-instructor slots, rider level rules, cancellations and reminders, then see where AI reminders add value.
- Ice Rink Software: Session Booking, Lessons and Party Enquiries — Set capacity from ice and hire stock, run lesson blocks with waitlists, take party enquiries to a deposit, and see where AI helps and where it must not.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Jackrabbit Swim product page and Jackrabbit Class feature menu (viewed October 2026); each vendor's own feature and pricing pages, to be checked on the day.