Why generative AI doesn’t fit into a standard in-housing playbook – yet
By Kimeko McCoy • August 20, 2025 •
Ivy Liu
The industry has moved functions in-house, establishing playbooks from everything from social to programmatic media buying. But brands are not using the same in-housing play for AI.
Some brands, like U.S. Bank, are taking a hybrid approach, while others, like Adobe, are more fluid, using AI tools both internally and in partnership with external partners. Mostly, brands are still testing and learning where AI can be leveraged in the marketing funnel, and what costs and where time can be saved.
“Over time, AI will likely shake up many brands’ attitudes about in-housing different parts of their advertising and marketing operations. But the ground beneath everyone’s feet needs to firm up a bit first,” said Max Willens, senior analyst at eMarketer.
Take U.S. Bank. The financial institution is using AI-powered marketing tools in everything from the customer experience to research analysis. But for heavier marketing efforts, like building synthetic audiences to serve as focus groups for real-time feedback, that’s handed off to Supergood, an ad agency owned by the holding company Supernatural AI.
There are a few reasons as to why a hybrid approach — mining out some AI functions while retaining others — is appealing, says Michael Lacorazza, CMO at U.S. Bank. On one hand, the brand doesn’t want to break up with its agency partners. On the other, it’s a heavy lift with tech that’s “not very user friendly,” making it cumbersome for a brand to adopt, he said. In large part, marketers are still trying to determine what efficiency looks like when it comes to AI-powered tools. Case in point, Digiday recently reported that some advertisers have started walking away from platforms AI solutions like PMax. Meanwhile, marketers are still trying to successfully use AI in brand building.
Other brands like Adobe and Unilever haven’t shied away from in-housing AI tools. Unilever has recently been building Beauty AI Studio, the bespoke, in-house system inside its beauty and well-being business, to make assets for paid social, programmatic display inventory and e-commerce usage.
As a B2B software company, Adobe has built out AI-powered tools, like Firefly, for external companies while also leveraging these tools in-house, said Stacy Martinet, Adobe’s vp of marketing and communications. She added, “We see AI as part of the ideation — a partner, not a replacement and creative, part of the production.”
The big selling point around in-housing marketing functions is retaining talent that works close to the brand that can turn around projects quickly, ultimately saving marketing dollars. For the past few years, brands have been in-housing creative teams, social teams and even media buying teams.
It’s not that brands don’t want to in-house AI technologies, according to Fred Schuster, CEO of InnerGroup, an in-house marketing operations company. It’s that internal agencies are often thought of as cost centers, lacking the budget or scale to build out and license AI tools. External partners, however, can invest in AI-powered tools, spreading those costs across clients or even reselling AI services to recoup costs, he said.
“If you’re external, you can make the argument that getting the money for AI to then sell it as a product to brands versus this is a tool a brand needs,” Schuster said.
Patrick Burgoyne, co-founder at In-House Agency Leaders Club, seconds Schuster’s point. In-house teams face barriers when it comes to adopting AI, including IP, copyright risks and other red tape. In Burgoyne’s opinion, it’s less likely AI results in increased headcount for in-house agency teams and more likely those teams will partner with external agencies that can deliver speed and efficiency. External agencies have already started moving to close the gap to make in-housing more accessible. Back in March, Jellyfish rolled out an AI-powered media in-housing platform, allowing advertisers to automate media research, evaluation, content strategy and campaign activation.
To both Schuster and Burgoyne’s point, it’s why U.S. Bank has a hybrid approach. “Once you build something, then you’ve got to maintain it, evolve it,” said Lacorazza. And maintenance means spending marketing dollars at a time when marketing budgets are flat and often accounting for 7% or less than their company’s revenue, according to Gartner’s 2025 CMO Spend Survey.
Still, as is with the in-house trend pendulum, expect it to swing in favor of more AI tools making their way to in-house agencies. As the tools get better and more democratized, brand marketers will likely take a second look at what they can do themselves, said Nicole Vinson, vp, global digital, media and omni-shopper experience at Kellanova.
“That’s the big question out there right now is, what is that right balance of in-house versus outsourced agency, content production model,” Vinson said.
Sam Bradley contributed to this report.
