Every school owner knows December is dead. Ask around at a seminar and you'll get it back as settled fact, usually with summer attached to it.
So we measured it. Three of the phrases an American parent actually types when they're looking for a school, pulled from Google's own advertiser data, month by month across the twelve months to August 2026.
"Martial arts near me" ran at 110,000 searches a month through September, October and November. December came in at 90,500. That's a dip, and it lands exactly on the twelve-month average Google reports for the term. It's also comfortably ahead of April, May, June and July, which sat at 74,000 in all four months.
"Karate classes near me" peaked at 40,500 in October and November, then fell to 18,100 in December. Less than half. That's the December everybody describes.
"Taekwondo classes near me" hit 9,900 in December. Highest month of the year. The low was July, at 4,400.
Same country, same season, same product, three phrases for it. One says December is the worst month on the calendar, one says it's the best month of the twelve, and one says the real trouble starts in April.
So the industry rule isn't a rule. It's one person's actual experience, generalised out loud at a seminar, and it may have nothing to do with your town or the families who walk past your door.
What that data is, and what it isn't
Before you do anything with those numbers, here's what they'll actually bear.
They're US national totals from Google's advertiser tooling, and Google reports them in rounded bands. 74,000, 90,500 and 110,000 aren't measurements, they're buckets. Anything smaller than a jump between buckets is invisible, so you can't read a 5% move off them.
They're Google, not Meta, and they're search intent rather than enrolments. Somebody typing "karate classes near me" in October is not the same thing as a body on your mat in October.
And they're national. A national average is the exact thing this article is telling you not to trust. Your county's school calendar, your local high school football season, whether your town empties out in July, whether there's a military base nearby with a posting cycle: none of that shows up in a US aggregate, and all of it shows up in your enrolment book.
So the numbers aren't your answer. They're the reason to stop borrowing somebody else's.
Read your own records instead
One school owner wrote to us in the middle of November, after a campaign had been running a few weeks, and the first thing he'd done was go and look:
"I just looked back over my enrollment this year and from January to October I had 41 students book a class or about 4 a month."
A school owner, in a message to Madscope. Source:
t07.png
Forty-one bookings over ten months. He didn't estimate that, he didn't feel it, he went and counted it. Which meant that when the question of December came up, he had something to compare against:
"If December being slow is several times slower than November, that's still several times more than I'm used to dealing with so I'd be okay with that."
Same owner, same message. Source:
t07.png
Notice what he's doing. He isn't arguing about whether December is slow. He's accepted that it might be, priced it, and decided he can live with the answer. You can only do that if you know your own baseline, and his whole baseline came from one afternoon with his own records.
He wasn't relaxed about the rest of it, either, and it's worth saying so:
"I appreciate the credit as seeing all the bookings plus the Meta ad spend adding up every day was getting a bit stressful, especially not having gotten any of the students actually paying anything in yet."
Same owner, same message. Source:
t07.png
That's what a busy month feels like from inside a small school, and it's the other half of planning for a quiet one.
Build the table in an hour
Open your billing or attendance system and pull every new student from the last three years. You want two columns and nothing else: name, and the month of their first class attended.
Not the month they enquired. Not the month their first payment cleared. The month they physically turned up, because that's the thing your business runs on and it's the only date that can't drift.
Then count by month, three years stacked. Twelve rows. That's the table.
Four things will try to lie to you in it, and all four are fixable.
1. Weeks open, not months. December isn't a month for most schools, it's two and a half weeks with a holiday break bolted on the end. August might be four weeks or it might be one, depending on your summer. Divide each month by the number of weeks you actually ran classes, or the shutdown will show up as a demand problem.
2. Your own behaviour is in there. This is the big one, and we'll come back to it.
3. One good month distorts everything. If April 2024 has 19 new students because a local school ran a demo day, that's an event, not a season. Look at each month across all three years before you believe it. A real seasonal pattern repeats. A good month doesn't.
4. Kids and adults are different businesses. If you run both, split them. Children's enrolments track the school calendar and the adults' don't, and averaged together they'll cancel each other out into a flat line that describes neither.
An hour's work, and you'll know the shape of your own year. No benchmark can hand you that.
The trap that ruins most of these tables
Here's the thing that makes a school's own history harder to read than it looks.
You've been told December is slow. So in December you stop advertising, you stop chasing enquiries, you run a reduced timetable, you're distracted by your own family, and you don't book a demo or a trial event because who would come. Then December is slow.
Your records can't tell those two things apart. A month where nobody wanted classes and a month where you quietly stopped offering them produce exactly the same row in the table.
We hear this from the other direction fairly often, from owners who found out what their own quiet season was made of once something was running through it. One wrote:
"Before I met you, we would rarely see any new students in December."
A school owner, in a message to Madscope. Source:
t29.png
Another, about two years in, put it this way about the holidays:
"Over here this time of year it's holiday season, and we almost never, in years past we didn't get new students this time of year. You could just forget about anybody visiting the dojo."
An owner speaking in a video testimonial on the Madscope channel.
Take both of those as what they are: two owners, describing their own schools, both of them clients of ours, which is not a neutral sample and we're not going to pretend it is. Neither one proves your December would behave the same way.
What they do establish is the weaker and more useful point. Some of what a school calls its slow season is a fact about the market, and some of it is a fact about what the school does in those weeks. You genuinely cannot tell which is which from records built in years when you did nothing. That's not a reason to assume the season is fake. It's a reason to know that your data can't answer the question, and to stop treating it as though it can.
Master William Wendell, who runs Marietta Martial Arts, gave the clearest seasonal read we've heard from an owner, and it doesn't match the folk rule either. His slow stretch runs September through December, and he doesn't expect things to pick back up until January. Not a December problem. An autumn one. He says it himself in a video on our channel.
The other half of the season is what it costs to advertise into it
If you're running ads through a quiet stretch, there's a second thing moving that has nothing to do with martial arts.
Meta sells ad delivery by auction. Your ad competes against every other advertiser trying to reach the same people, and the winner is settled on bid, on how likely the system thinks the person is to act, and on ad quality. The practical consequence is worth stating plainly: your costs can rise without you changing anything. Not your budget, not your targeting, not your creative. Somebody else showed up and bid.
In the fourth quarter, a very large somebody else shows up. Retail and e-commerce advertisers put their year's money into October, November and December, and pricing pressure concentrates around Black Friday and Cyber Monday. You're a martial arts school in one town, bidding for the attention of parents who are, that week, also being bid for by every retailer in the country.
What we're not going to give you is a number for it. The published benchmark data on Meta ad pricing moves in different directions in different years, and a percentage from somebody else's account, in somebody else's category, with somebody else's creative, would be worse than no number at all. The mechanism is solid. The size of it in your account, this year, is measurable only in your account, this year.
The point stands anyway. A quiet season and an expensive auction can land in the same weeks, and that combination is exactly why "we'll just go dark until January" feels so obviously right. It's also why going dark is worth thinking about harder than most owners do.
Four things to do with the quiet weeks
Turn the budget down instead of off. One owner handled a period of being full like this:
"I have requested to reduce my Facebook spend to the minimum required. Please let me know what that amount will be and we anticipate increasing it again by early December."
A school owner, in a message to Madscope. Source:
t47.png
Minimum, with a date to come back. Not zero, and not open-ended. A campaign that's been running has learned something about who responds in your area, and the restart in January is a cold start you paid for twice.
Book the season before it arrives, not during it. The lead time on this is longer than it feels. Another owner, planning a year ahead:
"I think we need to start advertising Summer Camp again around the middle of February. Please run it through the end of May... Then, we would need to start advertising next years After School Program around May 1. Let's run it through the end of August."
A school owner, in a message to Madscope. Source:
t50.png
Summer camp marketing starting in February. After-school marketing starting in May. Parents make those decisions months before the thing happens, and a school that starts advertising summer camp in June is selling into a decision that was already taken.
Compare that with an owner heading into his first summer with no history to work from:
"I would like to hear your thoughts on a different summer promotion or offer we can do? This is our first summer being open, and I hear it can be difficult to sign up new members."
A school owner, in a message to Madscope. Source:
t22.png
"I hear it can be difficult" is the whole problem in five words. Second summer, he'll have his own numbers and he won't have to go on hearsay.
Know what happens to the people already in the pipeline. This is the operational trap in going dark, and the same owner from earlier asked about it directly, before it bit him:
"A question on pausing, can we just stop the ads and hunting for new leads and just process the ones already in-system or does pausing mean the 19 leads would be in limbo? Or would they be handed off to me or... how would that work?"
Same owner. Source:
t07.png
Nineteen people had put their hands up and hadn't yet walked in. Pausing the ads and pausing the follow-up are two different decisions, and plenty of systems treat them as one. Ask the question before you flip the switch, not after.
Spend the quiet weeks on the work you can't do when you're full. The quiet stretch is the only time of year you can actually rebuild your follow-up process, sit in on your own trial classes, or work the list of people who already trained with you and stopped. That last one is the cheapest enrolment source you have and it takes evenings you don't have in September: how to work your former-student list properly. And if your trial-to-enrolment conversion has been bothering you, a quiet January is when you can actually watch where the other five trials went instead of firefighting.
Where we sit in this
We get paid when a student physically walks into your school, so a genuinely dead month costs us more than it costs you. That's the bias to read this article through, and it cuts the honest way here: we'd rather know which of your months are really dead than sell you advertising into weeks that can't produce anybody.
That's also why we're not handing you an industry seasonality chart. We haven't measured one, nobody's average predicts your town, and the three search terms at the top of this page are the proof that the averages don't even agree with each other.
Build your own twelve rows. If they show you a real season, plan the year around it and spend your quiet weeks on the jobs that need quiet weeks. If they show you a season made mostly of your own December habits, that's a much more interesting finding, and it's worth a conversation. Book a 15-minute call and bring the table. We'll ask what your worst month actually looks like, because it tells us more than any question on a form.
The one-hour version
- Export every new student from the last three years, with the month of their first class attended. Not enquiry date, not first payment.
- Count by month, three years stacked, and divide each month by the weeks you were actually open.
- Split kids from adults if you run both, or the two patterns will cancel out.
- Mark every month where you weren't advertising or weren't chasing enquiries. Those rows measure your behaviour, not your market, and you need to know which is which.
- Work backwards from next year's seasons and write the start dates in a calendar now. Summer camp sells in February. After-school sells in May.
The search volumes at the top of this page were measured on 16 September 2026 through the Google Ads search-volume API for the United States in English, covering the twelve months to August 2026, and are reproduced as returned. Google reports these in rounded bands, they are national aggregates, and they measure searches rather than enrolments; all three limits are stated in the body because they matter to how much weight the figures will carry. No industry seasonality benchmark, average or survey result appears anywhere on this page, because we have not measured one. Client quotes are reproduced as written, from messages sent to Madscope by the owners concerned, and from a video testimonial published on our own channel. Source images: t07.png, t22.png, t29.png, t47.png and t50.png. Nothing has been rounded, tidied or projected forward, and no owner's own caveat has been removed. The description of Meta's ad auction and of fourth-quarter competition in it was cross-checked against Meta's own published account of how the auction works; no percentage figure for holiday ad cost increases is stated, and no claim is made about January pricing, because the published benchmark data moves in different directions in different years and we would not be able to stand behind a number.