The honest answer is that there's no right percentage, and the rule of thumb you've been handed is answering a question you didn't ask. Five percent of revenue, ten percent of revenue, pick a number. None of them know anything about your school.
A percentage rule tells you what a business your size typically spends. It tells you nothing about whether your timetable can hold what it brings in, and nothing about whether your cash flow can carry it while it does. Those are the two things you actually need to know before you set a daily budget, and a school doing $30,000 a month with no free slots on the timetable and a school doing $30,000 a month with half-empty afternoon classes shouldn't be spending the same amount. One of them has somewhere to put new students.
Here is a more useful frame. Your ad budget isn't a number you choose. It's a ceiling you run into, and you run into whichever of these three is lowest:
- The cash you can float between paying the ad platform and the new students paying you.
- The space on your mat and the instructor hours to cover it.
- The spend you can hold steady for a full quarter without second-guessing it.
Find all three. The lowest one is your budget this month. That's the whole method, and the rest of this article is how to work each one out.
First, the floor
Before the ceilings, there's a floor, and plenty of schools are spending under it.
A budget too small to produce a readable result is worse than no budget, because you pay for it and learn nothing. If you spend enough to get four enquiries in a month, and one of them books, and they don't show, you now have a month of data that means nothing. Four is noise. You will conclude "ads don't work for my school" on evidence that couldn't have shown you anything either way.
The floor is the smallest spend that produces enough enrolments for the average to mean something. You can't know that figure in advance, which is uncomfortable, but you can bracket it: if your spend in a month can't plausibly produce double figures of enquiries, you're buying anecdotes, not information. Either commit to a number that can teach you something over eight to twelve weeks, or don't start. The half-hearted version costs money and settles nothing.
Ceiling one: the cash-flow gap
This is the constraint nobody warns you about, and it's usually the one that actually bites.
You pay Meta today. The family who clicked today books for next Tuesday. They show up Tuesday, take the trial, take a week to decide, enrol the following Monday, and their first payment lands some time after that. Meanwhile you've paid the platform every single day in between.
That gap is where growing schools get frightened. One owner wrote to us in the middle of an unusually good month, and the part worth reading isn't the good news. Watching "all the bookings plus the Meta ad spend adding up every day", he told us, "was getting a bit stressful, especially not having gotten any of the students actually paying anything in yet."
Read that carefully, because everything in it was going right. In the same message he'd worked out his own enrolment history: "from January to October I had 41 students book a class or about 4 a month. In November you've brought in 22 for just the first half of the month." The marketing was doing what he'd hired it to do. He was still stressed, and he was right to be, because the spend was daily and the revenue wasn't there yet.
He wasn't looking at a profit-and-loss problem. On paper those students were worth a great deal more than the ads had cost. He was looking at a timing problem, which is a different thing and can close a business that's technically making money.
So work out your own gap. From the day you spend a dollar to the day the student it bought makes their first payment, how many weeks is that in your school? Four? Six? Whatever it is, multiply your proposed daily budget by it. That's the amount of cash you need to be able to be out of pocket by, comfortably, before a single new student has paid you anything.
If that number frightens you, your budget is above this ceiling. Lower it until it doesn't. You can raise it out of the revenue once the first cohort starts paying, which is precisely how the owner above should have been told to start.
One more reason to set it at a level you can hold: stopping and restarting costs you momentum and it costs you the leads already in the system, and it's entirely avoidable by setting the number at what you can sustain for a full quarter rather than what looks affordable on the first of the month.
Ceiling two: the space on your mat
The second ceiling has nothing to do with money, and schools discover it with a jolt. One owner asked us to turn his ad spend down, not off, for one reason: "We have hardly any spaces left in our current classes!"
That's not a complaint, and it's not a failure. That's an owner reading his own timetable correctly and adjusting the tap. He goes on to say he's opening new classes at a new location the following month and will want separate ads for them, which is exactly the right sequence. Capacity first, spend second.
Another put a hard cap on it:
"I cant believe I'm saying this but can we maybe put cap limit of 5 per day and then I'll keep an eye on anyone who cant find a slot and do some manual bookings based on availability? Im not saying this because want to scale down, I just need to find more closers to keep up with what your team is doing!"
Note what he's capping. Not money. Arrivals. He's worked out that his binding constraint is the number of people who can be properly handled in a day, and he's throttling the input to match it rather than letting the overflow turn into bad first impressions.
This is the part owners get backwards. They set the budget from the bank balance and then discover the consequences on the mat. Do it the other way. Count the genuinely available places in your current timetable this month. Not theoretical capacity, not what you could take if you added a class you haven't staffed. Actual seats, with an instructor already assigned. Then work backwards through your own conversion rates: if it takes roughly four enquiries to produce a show and roughly two shows to produce an enrolment, then twelve available places means you want somewhere near a hundred enquiries a month, and no more. Spending for two hundred buys you families you can't serve, which is worse than not having them.
If you don't know your own ratios, that's the first thing to fix, and it's covered in more detail in what to do in the twenty minutes after a trial class ends.
Ceiling three: the spend you can hold steady
The third number is psychological and it matters more than either of the others, because it determines whether you can leave the campaign running long enough to do its job.
Ask yourself: could I hold this spend every month for three months without checking the account at eleven at night?
If not, the number is too high. Not because it won't work. It's because you won't be able to leave it alone. You'll change it after a quiet week, pull it after five, and never see what a campaign left to run would have done.
Pick the number you can commit to for a full quarter without flinching. That's the number you'll actually hold steady for long enough to read the result.
How to raise it
Once the money starts coming back, the budget question changes shape. It stops being "what can I risk" and becomes "what's the return on the next dollar", which is a far nicer problem.
One owner asked to raise his ad budget in January, and the reason he gave was a single line: "Dec. was a great month". Look at what the increase sits on: a month he already measured. He isn't raising the number because it's January and he feels optimistic. He's raising it because December told him something, and he's extending what December did.
That's the only defensible way to scale an ad budget. Raise it after a measured result, in increments you could reverse, and check the cost per enrolment after each step rather than after each quarter. That figure is how you know each step up is doing what the last one did.
And the way to know whether the figure is any good is to hold it against what a student is worth to you over their whole stay, not against your monthly fee. If you haven't done that calculation, do it before you touch the budget. Here's how to work out your own number. Hold the cost per enrolment against the total that student pays you before they leave, not against one month's $150 fee. The test is the plain one: once you're spending less to get a student than you make from that student, you're profitable, and you can do that all day.
The budget isn't only a size, it's a choice of where
One more from the wall, because it reframes the whole question:
"Hi :-) I think I need to pause the karate as everything is going great. We have lots of busyness, but I need to move the advertising budget to Crossfit."
He hasn't cut his marketing. He's moved it. The karate programme is full, so the money goes to the programme that isn't.
Most owners treat the ad budget as one number attached to the school. It's more useful as a number attached to a programme, reallocated when one fills up. The right question isn't "how much should I spend on ads". It's "which of my programmes has empty places, and what will it cost to fill them". The answer will be different for your kids' classes, your adult classes and your second location, because they've got different capacity, different value per student and different competition.
What to do this week
- Work out your cash-flow gap. Days from ad spend to the new student's first payment. Multiply by your daily budget. Can you carry that? If not, that's your ceiling.
- Count actual available places in your existing timetable, with instructors assigned. Convert backwards through your own show and close rates to an enquiry target.
- Name the spend you could hold steady for three months without second-guessing it. Divide by ninety.
- Take the lowest of the three. That's the budget. Run it for a full quarter without touching it.
- Then move it on evidence, in steps you could reverse, watching cost per enrolment after each one.
Nobody in this can tell you your number, including us, because three of the five inputs are facts about your building and your bank account. But the exercise takes an afternoon and it replaces a guess you'll second-guess every week with a figure you can defend.
If you'd rather your marketing company was only paid when a student actually walked through your door, that's how we're paid. The arithmetic above still applies; it just changes which risks sit on which side of the table. How the different ways of paying a marketing company compare is worked through in what each marketing billing model pays someone to do.