Mid-Sized Agencies Use AI To Build Tools Once Reserved For The Largest Holding Companies

Mid-sized agency groups can now use AI to build tools that once only the largest holding companies could afford. Media mix models are one example. They show advertisers how moving budget between channels will affect results. Holding companies have typically spent months and a large budget developing a single model. A mid-sized group can create its own version and run it for a fraction of that cost.

Tim Ringel is the Founder and Global CEO of Meet The People, an independent advertising group that has grown through acquisitions since 2021. A serial entrepreneur, Ringel has built multiple international agency groups and previously ran Spring Studios and Reprise Digital, IPG Mediabrands’ global digital agency network. He has managed advertising businesses across EMEA, APAC, LATAM and the U.S.

“If you’re an independent agency, you’re going to have to invest in AI. It’s like the switch from pen and paper to the typewriter. You can’t avoid it,” Ringel says. At Meet The People, savings from AI go toward building new capabilities. The group spends between $1 million and $2 million a year on its AI tool suite and has built eight proprietary tools, including its own media mix model.

Modeling on demand

Clients of every size want a media mix model, whether their budget is $1 million or $100 million. The models let marketers test what happens if they add $500,000 or $10 million in spend and what results to expect against their business goals. Ringel puts the typical holding-company price at about six months of work and around half a million dollars per model.

Meet The People launched its own model for digital channels in February and added traditional channels in November, combining both into one model. Results still depend on having enough fresh data, so the team runs scenario analysis about once a month. “Before AI, there was no way we would have been able to build our own MMM in that quality within a year. Today we can run MMMs at any time of day, and it costs us a few bucks in tokens,” Ringel notes.

Connecting every touchpoint

Buying a chocolate bar can involve 30 to 40 touchpoints between a customer and the brand. The customer might see an ad, order the product online, pick it up in a store, eat it and then buy it again. Each interaction produces data in a different place, such as legacy IT systems, CRM platforms or retail partners that may never share who bought the product. Pulling that data together used to require a team of data scientists and researchers.

“Those touchpoints were pretty much impossible to measure consistently. Now, thanks to AI, we can make sense of it in a heartbeat,” Ringel says. Connected data also changes how many partners a brand needs. A client that once hired several agencies to take a brand idea through to a TikTok execution can now work with one group that sees the full path from first impression to repeat purchase.

That shift makes room for a newer kind of agency group. Brands have long relied on two types of agencies. Large holding companies run much like the public companies they serve, with quarterly targets and client relationships that often go through procurement. Independent agencies make up the rest of the market. There are roughly 20,000 in the U.S. alone, ranging from two-person shops to firms of a few hundred people.

“There’s a new middle layer that’s trying to fill the vacuum between the holding companies with 100,000 people and the small independents that do one thing really well,” Ringel says. These groups mostly serve mid-market brands and work with clients over many years across several services. Handling both creative and media also gives these groups more data. A media agency that doesn’t make its clients’ ads has less information to feed into its models.

The 4As, the trade association for U.S. advertising agencies, sees the same trend. Its Look Ahead 2026 report describes two common directions for agencies. Some build their businesses around proprietary technology and data, while others focus on strategy and creative work. The report also points to a middle path that many independent and newer agency models are pursuing, with technology helping them grow quickly.

What acquirers want

AI is changing what makes an agency worth buying, partly because so many agencies are still early in adopting it. When the Interactive Advertising Bureau surveyed more than 500 professionals across brands, agencies and publishers for its State of Data 2025 report, only 30% had fully integrated AI across their media campaigns. Half of the industry had no strategic roadmap for AI. For groups that have already built their own technology, that gap matters less when they evaluate an acquisition. They can supply the tools and focus on an agency’s talent and client relationships.

Agencies without their own AI setup can also be easier to integrate, since there are no competing systems to merge. Ringel expects more clients to want agency partners in their own regions, which makes strong local agencies attractive targets as well.

Smaller agencies face the same question from the other side. When Meet The People acquired The LOOMIS Agency, a Dallas-based challenger brand specialist, this summer, LOOMIS hadn’t yet built out its own AI tooling. It now uses the group’s tools for its clients. “If you’re not joining a larger or mid-sized group that makes that tech investment, it’s extremely difficult, because these things are going to become table stakes,” Ringel says.

Walled gardens need guides

Over the past few decades, the media market opened up and is now splitting apart again. Before the internet, placing an ad often depended on knowing someone at the network or publisher. Programmatic buying removed those barriers and let any advertiser reach almost any publisher through one system. That simplicity is fading as TikTok, Amazon, Google, Meta and ChatGPT each run their own closed ecosystem, with separate buying tools and data.

For clients, every new platform brings new questions. “The more walled gardens you have and the more fragmentation, the more you need an agency to help you navigate. Every client calls and asks, ‘Could I run ads on ChatGPT?’” Ringel says. Answering that question well takes people who understand how each platform works and where it fits in a client’s overall plan.

More platforms and more automation are changing the skills agencies need. As AI takes over more specialized tasks, deep knowledge of a single trade or data source counts for less on its own. “People who are specialists today should learn to be a bit more of an architect within their subsection, because that’s what clients are going to need more,” Ringel says.

Building that talent from the ground up is its own challenge. When entry-level tasks are automated and teams work remotely, new hires have fewer chances to learn a craft like copywriting or graphic design from the colleagues around them. Even as AI raises how much revenue each employee can support, agencies still need to hire and train. “You can jump on the bandwagon of AI, but you still need people to implement it, make sense of it and put it into a strategy that works for you,” Ringel says.