"90% of gym websites look like they were designed by the same e-mail harvesting marketing company."
A student wrote that on Reddit about the schools he'd looked at. Four years later, in a different thread, another one: "It's like they all went to the same sketchy sales seminar in 1990."
Neither of them was talking about ads. But they put their finger on the thing that should worry you before you spend money on production. The shinier you make it, the more it looks like everybody else's.
Which is awkward, because the question owners ask us most often about creative is whether they should go and get some video done.
Should my martial arts school use video or image ads?
Run both.
Video isn't better than image in and of itself. That's not how advertising works. The question is never "do video ads perform better". It's "which ad is performing, the video or the image". Those are different questions and only the second one has an answer.
So we run both across the schools we work with, and an ad gets switched off for one reason only: it isn't getting students. The one that brings families in stays on. That's the whole decision, and your opinion about which looks better isn't part of it.
Neither is ours. Data is money in advertising, and decisions should be data driven, not opinion driven. Test all kinds of designs and copy, and let the data tell you what the best performer is.
Why do ugly ads convert better?
"Ugly converts" is an old advertising saying and it keeps playing out. The most polished, professional-looking ads often underperform against less beautiful ones.
Why, probably. A plain ad stands out in a feed full of polished ones. And people are often less forgiving towards large businesses and prefer smaller mom-and-pop places, so a slick ad can read as a chain when what the parent actually wants is the local school where the instructor knows her kid's name.
That's the same instinct behind those two Reddit comments. The template is recognisable from the outside. Your phone photo of a real Tuesday beginners class isn't.
None of which means ugly wins. It means the photo already on your phone deserves a two-week test before you decide it isn't good enough, because you're not the one it has to work on.
Do I need a professional video to run video ads?
No. And if you're holding off on advertising until the shoot is booked, you're waiting for nothing.
If you already have footage, send it. Anything from a real class. A grading, a line of white belts, a kid's first break. We'll cut it. If you'd rather film something yourself, we'll give you the script. If a school isn't going to film anything, we make our own, and that has never been the thing that held a campaign up.
If you have no footage at all, Meta will make the video for you. Inside Advantage+ creative in Ads Manager there's now a feature called video generation. Meta describes it as a generative-AI feature "designed to help you transform multiple, single images into a dynamic multi-clip video, with text overlays and music". So a handful of class photos become a video ad without anybody picking up a camera. It's still rolling out, so check your own account for it. Either way, it's free to test before you pay a videographer.
What is the ad actually being asked to do?
On Facebook the headline and the creative are usually the most attention-grabbing parts of an ad, so use them deliberately. Getting noticed is only the first half, though. Once you have her attention you have to make the offer effortless to understand, because a confused mind doesn't buy.
That's a copy problem far more often than a format problem. The ad that's failing is usually failing because it lists virtues, discipline and respect and confidence, that every school in your town also lists, in a feed where that parent has already scrolled past four of them this week. We've written separately about what to call your programme in the ad, and about which age bracket to point it at. Both will move your numbers further than swapping a photo for a clip.
How do I tell which one is working?
By who walks in, not by what a lead cost you.
Pay attention to what the targeting and the ads are getting you right down the chain: leads, bookings, students showing up for trial classes, enrolments. Knowing all of that downstream data, and not just clicks and leads, is what lets you make the call.
An ad can have a very low lead cost and produce nobody. That one isn't communicating what you think it is, and it probably wants cutting. An ad with a dearer lead that turns into students is the one to keep. Judge on the overall cost of getting a student to show up and enrol.
That's how we treat creative across every account we run. Call it survival of the fittest. An ad or a target that isn't getting students doesn't survive, and the ones that do get reproduced into more accounts. We watch it constantly, because we're not paid monthly. We're paid when a student walks in, so a pretty ad that doesn't produce one costs us too.
Here's the version you can run yourself this week.
The two-week creative card
Run both ads at the same time, in the same campaign, to the same audience, with the same offer and the same text. The only difference is the creative. Then fill this in at the end of week two.
| Image ad | Video ad | |
|---|---|---|
| Leads | ||
| Booked a class | ||
| Turned up | ||
| Enrolled |
Read it from the bottom up. The bottom row decides it. If both bottom rows are filling, you've got two winners. Leave them both running and go and write a third ad. If one bottom row is empty, that's the ad to cut, whatever its lead cost said.
If leads are landing on both and the bottom two rows are empty, the creative isn't your answer. That gap sits between the lead arriving and somebody replying to it. Text them inside fifteen minutes, from the number that's already messaged them, and watch those two rows move. Here's the full version of that.
Why did my ad stop working after a few weeks?
Because Meta ads fatigue. The parents in your radius have now seen that ad several times and have stopped seeing it. This is normal, and it isn't a sign the campaign was wrong.
What it needs is new ads, new hooks, new angles, ideally made before the current one tires, so there's always something fresh going into rotation. That's one of the real arguments for keeping a mix of image and video on hand. It's two sets of raw material to make the next round out of instead of one.
If you're putting video on TikTok as well, budget for the same thing at a faster pace, because it eats content. One trick makes that survivable: record several different intros and several different endings, then edit the same middle together in different combinations. An afternoon of filming becomes a fortnight of posts.
The thing to do on Monday
Before you pay anyone to film anything, find the best photo you already have of a real class. Phone photo, parent's photo, doesn't matter. Run it as an image ad beside your current video for two weeks. Same audience, same offer, same words. Fill in the card above.
Then let the bottom row decide, not your eye for a nice picture.
If you'd rather somebody else ran that test, that's the job. We find the families, get them booked and remind them to come. You welcome them, run a great first class, and enrol them. We build the creative, we run both formats, and we keep whichever one puts people on your mat. Across 400+ schools, which is how we already know which hooks are working somewhere else this month.
No retainer, no contract, and we're paid when a student walks through your door. If you want to talk it through, book a call and bring your timetable and the number of spaces you can genuinely take this month. If we don't think we can fill them, we'll say so on the call.
For the wider setup around the creative, the objective and the follow-up and the mistakes that cost most, start with what actually works in martial arts Facebook ads.
Questions owners ask
Should my martial arts school use video or image ads on Facebook?
Both, at the same time. Video isn't better than image in and of itself. That's not how advertising works. The real question is which ad is performing, the video or the image, and the only way to know is to run them together and look at who booked, showed and enrolled. We keep whichever one brings families in and switch off whichever one doesn't.
Do I need a professional video shoot for my school's ads?
No, and it isn't where we'd send your money first. If you already have footage from a class, send it. If you don't, Meta's Ads Manager now has a video generation feature inside Advantage+ creative that builds a short multi-clip video out of your still photos, with text and music. Test what you already own before you pay anybody to film.
Why do ugly ads convert better?
Polished, professional-looking ads often underperform against plainer ones. The likely reasons: a plain ad stands out in a feed full of polished ones, and a slick ad reads as a large business, while people tend to be more forgiving towards a small local school. It isn't a rule. It's a reason to test the photo on your phone before you write it off.
Why did my Facebook ad stop working after a few weeks?
Meta ads fatigue. The same people in your area have now seen the same ad several times and have stopped noticing it. The fix isn't a bigger budget behind the worn-out ad. It's new ads, new hooks and new angles, queued up before the current one tires, so there's always something fresh in rotation.