AGS MAGAZINE: Creative is the new media buying

There is a bigger shift happening in performance marketing and creative today that many might not notice or act on. Iveta Makedonska, Founder & CEO of MakeItClick, talks about why creative has become the new media buying, why most brands don’t have a content problem but a testing problem, and how AI won’t replace creative teams but will completely change the way they work.

When creative became targeting

A few years ago, you only needed a great media buyer, a strong offer and a healthy ad account to scale your ads. Today, you can still hit a ceiling if you can’t continuously produce new winning creatives. This is because media buying has become highly automated. Meta can optimise an ad’s distribution, but creative provides the algorithm with more signals about who your audience is and what resonates with them.

There are still many people sleeping on this shift and persisting with their old ways, only to see their campaigns struggle over time.

“That’s when it became very clear to me that the biggest leverage was shifting towards creative. The brands that could consistently find new angles, new audiences, and new ways of communicating the product had much more room to scale,” says Makedonska.

That makes creative one of the biggest differentiators between campaigns that keep scaling and those that start struggling. Makedonska sees this happen with her clients. Two brands can be running very similar Meta setups, targeting broadly and spending at a comparable level. One starts struggling with rising CPAs and assumes there’s a media buying problem. The other keeps introducing genuinely new creative hypotheses, such as different personas, problems, use cases, and formats, and continues to find pockets of growth.

The difference isn’t that one brand has discovered some secret targeting strategy. It’s that they’re giving the algorithm more opportunities to find demand.

For Makedonska, that’s an important distinction: “Creative has become a form of targeting.”

The end of one-hit-wonder ads

Every agency and brand dreams of that viral ad, hoping a single hit will put the brand in the spotlight and drive revenue. But for Makedonska, relying on one viral ad has its limits. Instead, she’s in favour of building a creative engine that continuously tests, learns and improves.

Day-to-day, a creative engine is a combination of team structure, process and software. But the process is the most important. It starts with research, which Makedonska sees as the foundation of the creative engine. And that research never really stops. Creative strategists should be researching, monitoring their accounts, and continuously learning from performance.

From there comes creator selection and understanding which creators fit the audience and concept. Only then do you write the script, produce the ad and move into editing and testing.

You need a system in which research generates hypotheses, hypotheses become concepts, concepts are developed and tested, performance yields insights, and those insights feed directly back into the next round. So instead of saying, “This ad won, let’s make ten more versions,” Makedonska asks why it won. Was it the persona, the hook, the use case, the creator, the format or the level of awareness? From there, the team decides what deserves iteration and what completely new hypotheses should be explored.

This also means creative strategists, editors, creators and the performance team shouldn’t operate in separate bubbles. There needs to be a continuous feedback loop between what you’re producing and what the market is telling you.

The goal, according to Makedonska, isn’t to find one winning ad. It’s to build a system that can repeatedly find winners.

Brands have a testing problem

Testing ads has also evolved. Gone are the days when you could test three concepts with ten variations and call it a testing strategy. A brand that thinks it’s testing might say, “We launched 30 new ads this month.” But look closer, and they may essentially be three ideas with ten different hooks.

That’s content volume, but it isn’t necessarily testing volume. If five ads look different but communicate the same idea to the same person, you haven’t really created five meaningful tests.

Today, testing has to go deeper. If Makedonska were testing 30 ads, she would develop 10 different concepts, each with 3 variations, while making those concepts distinct enough to learn from them. That gives the team more data and gives Meta enough variety to determine what resonates.

Maybe you’re testing whether a completely different persona responds to the product. Maybe it’s a new use case. Maybe you’re moving from problem-aware messaging to unaware messaging. Or maybe you’re testing whether education outperforms aspiration.

There are also different elements within those concepts to test. For video, that could mean two-minute, one-minute and 30-second versions, along with different hooks, editing styles and caption formats.

“Real testing means you can clearly say what you’re trying to learn,” says Makedonska.

What AI leaves to humans

Makedonska believes AI won’t replace creative teams, but it will completely change the way they work. One of the biggest changes she has already seen is in research.

There used to be a huge amount of manual work involved in going through reviews, comments, competitor ads, customer feedback and other sources to identify patterns. AI can now help her team process and organise an enormous amount of that information much faster.

It can also analyse creative performance in more depth, picking up patterns across video length, hooks, formats, creators and other elements that would previously take someone much longer to find manually.

But this hasn’t made strategy less important. If AI gives everyone the ability to generate 100 hooks in seconds, generating hooks itself becomes less valuable. Knowing which hook is actually worth making becomes more important.

So the time saved is reinvested in interpreting research, developing stronger hypotheses, analysing performance, and thinking more deeply about what to test next.

For Makedonska, “AI removes a lot of the labour around creativity, but not necessarily the thinking.”

The reality is that AI will keep making ads cheaper and easier to produce at scale. That also means more brands will be producing at the same volume. Quality becomes more important because simply producing more ads is no longer enough to stand out.

Where she’d put her money

If Makedonska had to bet on what separates the brands still scaling in three years from the ones that stall out, it wouldn’t come down to who produces the most ads or has the biggest team.

She would put her money on the brands that can learn faster than their competitors. These are the teams that can take customer insight and performance data, turn them into a hypothesis, get that hypothesis into the market quickly, understand what happened and immediately use what they learned to make the next decision.

Platforms will change. Formats will change. AI will change how content gets produced. Brands that can adapt to those changes and turn what they learn into better creative will have an advantage. For Makedonska, speed of learning will matter more than simply speed of production.

That mindset also changed how she looks at failure. When Makedonska was starting, she says she was a perfectionist who saw mistakes as unacceptable. Over time, she learned that testing is bound to fail. Finding winning use cases, personas, and concepts also means accepting that some tests won’t work.

Looking back at everything she learned scaling 100+ brands, Makedonska would tell her younger self to get obsessed with understanding why something works, not just whether it works.

That mindset has influenced how she built MakeItClick more than anything else.

This article was originally published in AGS Magazine. Explore the latest issue for more in-depth interviews, straightforward analysis and informed perspectives from across the affiliate and performance marketing industry.