For all the talk about AI transforming marketing, many brands still wait weeks just to push a single audience from a CRM into paid media – a lag that’s hard to square with the rise of autonomous AI agents.
“But that’s actually pretty normal in the industry,” said Rio Longacre, a partner and global ad tech lead at Credera, Omnicom’s consulting division.
In other words, good luck building agentic workflows for online advertising on dysfunctional plumbing. “It’s either just not going to happen,” Longacre said, “or is going to be really messy.”
To address this “really messy” reality, Credera has been testing new technology from infrastructure provider MadConnect that allows buyers to plug AI tools into existing ad tech and marketing platforms without having to set up every connection from scratch.
On Tuesday, MadConnect made that software, which it refers to as its “intelligent connectivity layer” (ICL) generally available to enterprise marketers and their agencies.
“Let the agents do all the innovative and fun stuff,” said MadConnect CEO Bob Walczak. “We just want to be the infrastructure, the connection point that they connect into.”
Plugging away
In practice, being the connection point means handling all of the unglamorous work of wiring CRMs, CDPs, clean rooms, DSPs and other systems together so AI agents can be useful across them instead of being trapped inside a single platform.
Under the hood, ICL is built to natively support Model Context Protocol (MCP), an open-source spec that standardizes how AI agents share context and interact across systems securely.
MCP defines how an agent talks to a data source or tool. But it doesn’t actually connect those systems to each other. That’s where ICL comes in, sort of like a switchboard behind the scenes.
If MCP is akin to a USB, think of ICL as the row of ports on a laptop.
“The promise of AI is that you don’t need engineers to do a lot of things, but the fact is that you still need technical acumen to use things like MCP servers, APIs, AdCP [and] UCP,” Longacre said. “This lightens the burden and gives you one place where you can pretty easily drag, drop and connect things.”
Pain in the pipes
Which is important, because the integration grind in ad tech is real, and it doesn’t go away just because AI agents are in the mix.
Dentsu, which has also been using ICL, maintains a long list of direct integrations on behalf of clients, including with publishers, ad tech platforms and mar tech tools. It’s necessary, but also a colossal pain in the rear, according to Gerry Bavaro, Dentsu’s chief solutions officer for data and tech.
“Lighting up a new connection always takes time: setting it up, maintaining it and then adding layers for AI and measurement,” Bavaro said. “There are a lot of potential choke points between signing a deal and having everything actually running.”
Getting data back from partners is another bottleneck. Dentsu now uses ICL’s preconfigured connections to pull in log-level data and data from conversion APIs rather than having to do a whole bunch of custom integrations.
Platform neutrality
But Walczak is careful to frame MadConnect as a neutral infrastructure provider and not just another middleman.
The company doesn’t own an identity graph, isn’t trying to be a CDP and doesn’t sit in the media supply chain, he said. It runs within a client’s existing stack, where it manages connections and never takes custody of data.
This zero-copy data governance approach aligns with the zeitgeist. Most large brands want composability and limited data movement for privacy and regulatory reasons, Longacre said.
“It’s nonnegotiable for a lot of big brands,” he added.
Once the governance and connectivity issues are taken care of, however, the use cases start to look a lot more like agentic marketing than basic automation.
For example, early adopters of ICL are testing agentic-style workflows, such as buyer and seller agents exchanging inventory and handling reporting, with MadConnect managing the connections into the ad server and DSPs.
Bavaro expects this kind of infrastructure to also impact how agencies get paid. Dentsu already has some clients on commercial models that tie fees to outcomes, including ROAS and cost-per-acquisition efficiencies. Smoother connectivity and better data access makes it easier to take on that risk.
The bigger question now, though, is whether this kind of infrastructure layer can move AI in ad tech out of the demo phase.
“That’s the problem with AI – how do you go from pilot to production?” Longacre said. “We think tools like this can help us bridge that gap and move a lot quicker.”
