AI has officially lost its mystique—it’s no longer a shiny perk, but a table-stakes expectation in media agency reviews. While brands might not call it a “requirement” in writing, agencies are treating it as such after watching clients deep-dive into tech demos and start poking holes in AI models like they’re shopping for a used car. Just a year ago, a flashy PowerPoint would’ve sufficed. Now, agencies like Horizon Media are pulling back the curtain on their platforms, even sharing SQL code and letting clients tweak definitions in real time. Don’t have a working AI model? Don’t even bother walking into the room.
The problem is, while agencies are expected to be AI fluent, brands don’t always know how to speak the language. RFIs ask vague questions like “How do you use AI?”, forcing agencies to both guess and educate. Case studies and plug-and-play demos have become critical, but leading the conversation is just as important. Agencies are seeing success not just by presenting the tech, but by reframing the assignment—decoding real business needs, steering marketers toward meaningful use cases, and in some cases, telling clients to ask better questions.
All of this is reshaping the way agencies pitch, price, and even get paid. Outcome-based compensation is inching in as AI shifts focus from hours to results. For now, success looks like guiding brands through both what AI can do and what it actually should do. As one exec put it: agencies aren’t just showing their tools—they’re helping clients understand the job to be done.
Full story – with additional detail on Horizon’s approach and examples from Monks, IPG Mediabrands and more – at AdAge.