Joshua Kalms Says Advertising Has a Judgment Problem

Joshua Kalms spends his working days looking at advertising across fintech and AI companies, and he sees an uncomfortable contradiction developing. Businesses can produce creative faster than ever, yet the ability to generate more advertisements has not necessarily made those advertisements more distinctive. As CEO of Multiply, a creative performance agency focused exclusively on those two sectors, Kalms believes the scarce resource is shifting from production capacity to creative judgment.

"Making another ad has become incredibly easy," Kalms says. "That changes where the difficult work sits. The question is no longer whether you can produce more. It is whether anyone has made a strong enough decision about what is actually worth producing."

Fintech companies often have complex products to explain while operating within significant regulatory constraints. AI companies face another problem as competitors make similar claims around speed, intelligence, efficiency, and cost. Through Multiply, Kalms works across numerous accounts in both categories, giving him a view of recurring creative problems rather than the experience of a single brand.

"When you are seeing the work across multiple companies in the same categories, patterns become easier to recognize," he says. "You start to see which problems belong to one business and which ones keep showing up in different forms."

One of those recurring problems, in Kalms' view, is volume being mistaken for variety. AI tools can reduce the time and effort required to produce creative, making it tempting for teams to respond by generating more versions. Kalms argues that greater output offers little advantage when those versions are built around the same familiar ideas.

His concern is especially pronounced in AI advertising, where he sees increasingly similar creative emerging within an already crowded category.

"There is an irony in AI companies using AI to make advertising that looks like every other AI company's advertising," Kalms says. "Production efficiency is useful, but efficiency does not give you a point of view. When everybody can access similar tools, the decisions around those tools become more important."

That is why Kalms pushes back against the assumption that more creative automatically means better performance. His argument is that judgment becomes more valuable as the cost and difficulty of production fall.

"If a competitor could arrive at essentially the same idea with a prompt, I would question how much strategy is really there," he says. "Technology can help execute an idea. It cannot make sameness distinctive just because you produced it faster."

Fintech creates a different kind of pressure. Kalms says customer acquisition costs in the category have risen by more than 40 percent since 2023, while changes following Apple's App Tracking Transparency have weakened some of the targeting capabilities advertisers previously relied on. He sees those shifts as part of the reason creative now carries more responsibility within the acquisition process.

"When targeting becomes less precise, the ad has to do more of the work," Kalms says. "You cannot rely on the platform to solve every audience problem for you. The creative itself has to make the right person recognize that the message is for them."

Regulation adds another layer, particularly for financial products. Kalms disagrees with the common assumption that compliance requirements inevitably make advertising less interesting. He sees constraints as a test of whether a creative team has found an actual idea or is relying too heavily on product claims.

"If the only interesting thing about an ad is the claim, then compliance can feel like the enemy," he says. "Once certain claims or numbers are off the table, you are forced to think harder about what the audience should remember. That can lead to better work because the idea has to carry more weight."

He believes the strongest teams learn to work within those boundaries rather than treating regulation as an excuse for forgettable creative. For Kalms, the challenge is to understand what cannot be said, then use that clarity to find another way into the message.

"Constraints narrow some options, but they can also remove lazy ones," he says. "You still have to make something people notice. The limitation does not remove that responsibility."

Kalms sees a related issue as AI companies move beyond digital acquisition and invest more heavily in television, video, and out-of-home advertising. He agrees with the broader move toward brand building, but he argues that changing channels does not solve a creative problem by itself.

"You can buy television and still say something nobody remembers," he says. "A bigger format does not rescue an ordinary idea. The company still has to give people something distinctive enough to recognize later."

That perspective also informs his skepticism toward the traditional division between brand and performance. Kalms does not see memorable creative and measurable results as competing goals. In his view, distinctive advertising can also perform better, which is why he rejects a strict division between brand work and performance work.

"The idea that brand work lives over here and performance work lives somewhere else has never made much sense to me," he says. "If an ad is distinctive and people remember it, that can help performance. You should not have to choose between being memorable and being effective."

That philosophy sits at the center of Kalms' broader view of where advertising is headed. As AI continues to compress the time required to produce creative, he expects the gap between producing work and making good decisions about that work to become more visible.

He also believes that difference will matter commercially. Companies with a clear point of view may be better positioned to stand out as more competitors gain access to the same production tools. Those that rely on easily reproduced creative risk making every advertising channel feel more expensive without solving the underlying problem.

"The tools are going to keep getting better," Kalms says. "That makes the human decision more exposed. When everyone can make something polished, the question becomes who has something worth noticing."