Test Online, Then Print: How to Split-Test Your Martial Arts Marketing

The way to split-test martial arts marketing is to test on Facebook first, find the winner with real data, and only then put it on door hangers, postcards, rack cards and banners. Online platforms can split-test in real time with enough volume to tell a real winner from luck. A local print run almost never can. So don’t force your online ads to match your print. Make your print match your proven online winner.

I call this Prove It, Then Print It. I’ve wanted to work this way for more than 40 years, and for most of that time I couldn’t.

Why couldn’t martial arts schools split-test in the print era?

Because you needed enormous volume to get an answer, and small schools never had it. Back in the early 1980s in Washington, D.C., I was practically living at the Library of Congress, reading every direct-response source I could find. The lesson in all of them was the same: A/B split-test everything. Run two versions, keep the winner as your “control,” and then try to beat it.

In the era of newspaper, TV and coupon mailings, that was much easier said than done. Later I tried to split-test direct mail with Tim Kovar and Keith Hafner, working with Jay Abraham, to get to a control piece. We never could tell the difference between one piece and another. We weren’t mailing a million pieces at a time, and the statistical relevance of what we were mailing just wasn’t very high.

Why can’t a small mailing pick a winner?

Because when responses are rare, random chance swamps the real difference between two pieces. Here’s a simple illustration. These are round numbers I picked to show the math, not results from any campaign.

Say you mail 5,000 of postcard A and 5,000 of postcard B. A gets 25 responses (0.5%) and B gets 30 (0.6%). B looks 20% better. But with counts that small, random variation is roughly the square root of the count: about 5 responses either way on each card, and about 7 on the gap between them. The 5-response gap is smaller than the noise. Mail the same two cards again next month and A could easily win.

How many would you need? Using the standard sample-size formula (95% confidence, 80% power), reliably detecting a real difference between a 0.5% and a 0.6% response rate takes roughly 86,000 pieces of each version. That’s over 170,000 pieces for one test. No local school mails that much, which is exactly why we never found our control.

Why can Facebook and Google pick a winner when print can’t?

Because they generate thousands of data points a day and measure them in real time. Every impression, click and form submission is counted, and the platforms can split your budget between versions automatically. As of September 2026, Meta offers A/B testing in Ads Manager and a Creative Testing option in the ad setup that splits budget evenly across two to five test ads so the algorithm doesn’t favor one too early.

Same illustration, online. Two ads each get 20,000 impressions. Ad A gets 200 clicks (1.0%) and ad B gets 260 (1.3%). Because the counts are much bigger, the noise on the gap is about 21 clicks while the gap itself is 60, almost three times the noise. That’s a real difference, and you can see it in days instead of waiting a month for the mail to come back.

One caution: clicks are only a first signal. Judge the final winner on leads, booked intros and enrollments. The strongest winner produces students, not just clicks.

What happens when you force your Facebook ads to match your print?

You throw away the one advantage digital gives you. On one of my coaching calls, a school had a local print advertiser running door hangers. The school asked the ad agency we work with to make its Facebook creative match the door hanger so everything looked uniform. Once the creative was matched, lead volume dropped off sharply. The agency told them straight: you have to diversify and test.

“If you handcuff them and try to match them to print, you’re going in the exact opposite direction.”

Print can’t tell you what works. Facebook can. So let Facebook lead and let print follow.

How does Prove It, Then Print It work step by step?

Step 1: Build a diverse creative portfolio online

Run several genuinely different ads at once, not five versions of the same picture. The agency recommends a mix:

• Short-form video: a real class, a real instructor, real energy.
• Static ads with real studio photos: your mats, your students, your building. Not stock.
• Holiday and seasonal angles: according to the agency, these are still the best performers.

Step 2: Let the platform split-test

Use the platform’s testing tools so each version gets a fair share of budget. Give it enough time and spend to build up real numbers. Don’t kill an ad after one day because you don’t like it.

Step 3: Pick the winner on enrollments, not likes

Follow each ad through to cost per lead, cost per booked intro and, above all, cost per enrolled student. The winner is the ad that fills your classes.

Step 4: Take the leap to print

Now take the winning headline and image offline: door hangers, rack cards, postcards, banners. You still can’t measure the print piece with statistical precision, but you don’t need to anymore. You’re printing something that already has real data behind it instead of a guess.

Step 5: Keep testing online

Don’t freeze your Facebook creative to match the print you just ordered. Keep the online tests running. When a new winner emerges, that becomes your next print run.

Is AI-generated ad creative helping or hurting?

Mostly hurting, when it’s used lazily. The agency’s observation is that AI-generated creative is making Meta ads all look the same. Their advice stuck with me: “It doesn’t have to look great, it just has to look different.” A real photo of a real kid in your real class stands out in a feed full of polished AI graphics that look alike.

My own gripe is clutter. When you ask ChatGPT for a graphic, it throws in a phone icon, a map icon, a globe icon and the other 22 pieces of garbage. I recently walked through a farmers’ market and saw banner after banner that was obviously made by ChatGPT. Every one was cluttered, and none of them grabbed me.

Here’s what a good ad needs:

• One powerful headline.
• One visual. For your kids’ program, a smiling kid in class. For your adult program, an adult working out.
• Nothing else fighting for attention.

If you use AI tools, direct them hard. Tell them exactly what you want and what you don’t. If the tool keeps adding clutter, screenshot its output, upload it, and say, “Don’t do this.” It learns fast when you show it the mistake.

My rule of thumb, which is my own estimate and not a measured statistic, is that on Facebook about 80% of an ad’s effectiveness comes from the visual, 15% from the headline, and 5% from everything else. That’s why the visual is the first thing to test and the first thing to carry into print.

What should you do this month?

• Audit your current Facebook creative. If every ad looks the same, or looks like your door hanger, you’re not testing.
• Launch at least three genuinely different ads: one short video, one real-photo static, and one seasonal angle.
• Track each one through to enrollments.
• Hold your next print order until you have a winner. Then print the winner.

For the rest of the system, see my martial arts school marketing hub.

FAQ

Can I A/B test postcards or door hangers for my martial arts school?

You can try, but a local run is almost never big enough to tell a real winner from chance. In my illustration, spotting a difference between a 0.5% and 0.6% response rate takes roughly 86,000 pieces per version. Test online instead and print the winner.

Should my Facebook ads match my print advertising?

No. Let your online testing decide the creative and make the print match the proven winner. Forcing online ads to match print takes away your ability to test.

Are AI-generated graphics good for martial arts ads?

Only if you direct them hard. Default AI graphics tend to be cluttered and look like everyone else’s. Use one strong headline, one real visual, and nothing else.

What kind of Facebook ad creative works best for martial arts schools?

Keep a diverse mix of short-form video, static ads with real studio photos, and holiday angles. The ad agency we work with reports that holiday angles are still the best. Your own tests should make the final call.

Your Next Step

If you want the complete system for filling your school, get my free book, Six Simple Steps to Add 100 Students, at FillYourSchool.com. And if you’d like my team to look at your numbers and show you where your enrollments are leaking, book your free Personal Evaluation (a $1,297 value) at martialartswealth.com/go/evaluation/.

About the Author

Grand Master Stephen Oliver, MBA, is a 10th Degree Black Belt and the Founder and CEO of Mile High Karate and Martial Arts Wealth Mastery, one of the martial arts industry’s leading coaching and consulting organizations for professional martial arts school owners, BJJ academies, and MMA gyms. A martial arts school owner since 1975 and business coach since 1985, Oliver has spent more than five decades building, operating, and advising successful martial arts schools. He also serves as CEO and Chairman of NAPMA (the National Association of Professional Martial Artists) and Publisher of Martial Arts Professional magazine. A Georgetown University cum laude graduate, he earned his Executive MBA through the Executive Program at the Daniels College of Business at the University of Denver. He was promoted to 10th Degree Black Belt in April 2026 and inducted into the Official Taekwondo Hall of Fame in August 2026. Learn more at MartialArtsWealth.com and StephenCOliver.com.