What AI disruption means for experimental ad budgets

Dollars shifting out of performance

Take performance marketing shop Markacy for example. Historically, the agency followed an 80-20 rule: 80% of spend goes to core media and 20% to experimenting. 

“There’s a little bit of a tailwind for a bigger expansion budget that might be getting closer to 25% or something like that,” said Christopher Jones, co-founder and co-CEO at Markacy.

Over the past year, the agency has seen clients more willing to shift dollars out of Meta and Google and into new channels, like CTV, YouTube, linear, direct mail, out of home and retail media networks. 

It’s a similar story at creative media agency Noble People. Dollars that were earmarked for platforms like Meta and Google are shifting to experimental, upper-funnel media or “the big stuff” as Nitin Sinha, head of media planning at Noble People, puts it. 

“We’re absolutely seeing clients more and more interested in experimentation. They want to create buzz, they want to create real high-impact [marketing],” Sinha said. He declined to outline specific budget figures.

Pawco pet food brand too has revamped its experimental ad budget, which now accounts for 15 to 20% of the pet brand’s marketing budget, which is relatively high, according to Ryan Bouton, vp of growth at Pawco. Specifically, Pawco saw a 10% increase to its budget in Q1 of this year just to account for new AI/AEO tactics, Bouton added. 

AI forcing test and learn strategies

Economic uncertainty and tariffs have been partly to blame for the experimental ad push. There’s also tech advances in channels like retail media and CTV – and the fact that brands need to be more exploratory to regain deteriorating reach on performance channels like Google and Meta. 

The primary driver, however, is AI forcing businesses to test new approaches to keep up with shifting consumer habits. Markacy clients are increasingly requesting AI search-related services, per the agency.

“The AI environment has just been..it’s scary for some, but it’s a tail-in in terms of testing, because there’s no other option. You have to lean into new things,” said Markacy’s Jones.

As OpenAI pilots ads within ChatGPT, brands like The Knot are funding ad pilots with the AI platform using test and learn dollars. 

For some marketers, experimental ad dollars are baked into the overall marketing budget. It makes sense as advertisers have come to expect uncertainty as the new norm. Meaning, ad spend strategy has been increasingly founded in marketers’ ability to navigate an unstable, fragile market.

Flexibility baked in

BEHR doesn’t break out experimental dollars. Instead, the paint brand has an annual planning process to understand where dollars work the hardest and shifts spend from there. Ogilvy North America has seen clients take a similar approach. 

While clients are significantly increasing standalone experimental budgets, there’s more collaboration to explore AI-driven opportunities without adding line items to the budget, Kaare Wesnaes, head of innovation at Ogilvy North America, told Digiday in an email. 

As performance channels erode, economic fragility continues and AI upends the digital landscape, marketers say test and learn dollars will continue to play a crucial role in marketing investment.

“Instead of trying to force this false sense of precision around attribution, companies — at least the good ones — are getting more comfortable looking at things holistically,” Jones said.