Technology
Basis Platform
Advertising Operating System
Communication Portal
Intelligent Collaboration
All Channel Activation
Activate Media on Every Channel
Consulting
Resources
Key Takeaways:
Few technologies have reshaped marketing as quickly as AI.
In the less than four years since ChatGPT’s public debut, AI has become increasingly embedded in many marketers’ workflows, influencing everything from media buying to creative development. Now, its role is expanding.
Agentic AI—autonomous systems that both generate outputs and execute decisions, often with less human review at each step—is moving deeper into advertising workflows. In fact, 77.7% of agency leaders plan to increase their AI investment over the next 12 months and 90.7% of industry professionals believe that AI will radically transform the industry within the next three to five years. Yet amidst this enthusiasm, teams also need to consider whether they have the data infrastructure needed to support this rapid pace of adoption.
When AI runs on fragmented or low-quality inputs, results are unreliable. Insights get blurred, personalization misses the mark, and recommendations fall flat. Such missteps have a direct cost, quickly compounding into wasted spend and weakened consumer trust. The organizations that thrive in the AI era will be those that invest in clean, unified, privacy-compliant first-party data—the foundation AI needs to deliver accurate, differentiated value.
From faster analysis to smarter targeting to more relevant creative and beyond, the potential for AI in marketing is significant. But the reality of implementing AI effectively is that its accuracy hinges on the quality of the data it consumes.
When data is inaccurate, siloed, or inaccessible, the risk of error increases significantly. Poor data muddies results and raises the odds of hallucinations, where AI generates outputs that appear credible but are instead fabricated. Because the technology mimics the information it’s trained on, gaps or inaccuracies in the data increase the likelihood of such mistakes. This is a significant problem, with a recent study finding that nearly half of marketers encounter AI inaccuracies several times a week.
Those hallucinations can look like real insights: an optimization tool shifting spend toward audiences built on incomplete signals, a personalization engine delivering irrelevant product recommendations with full confidence, or a dashboard surfacing “top-performing” keywords that don’t exist. These are the kinds of costly missteps weak data foundations can produce. Even small inaccuracies can snowball, feeding back into models and negatively shaping future decisions. And as AI moves from assisting to acting, the stakes grow even higher: An agent executing decisions autonomously removes the human checkpoint that might otherwise catch these errors before they compound into a much larger issue.
Agentic AI is top-of-mind for US ad buyers. Two-thirds say agentic AI ad buying and execution is an increased focus this year, and 84% cite media planning and buying recommendations as a current or likely use case. Yet, despite this interest, many are also hesitant: 40% of buyers report that understanding agentic ad buying and campaign execution is one of their greatest concerns or challenges at present.
Data is often a meaningful part of that hesitation. Understanding how an AI agent functions means understanding the data it acts on. Generative AI and agentic AI can both produce errors that are difficult to catch—such as biased outputs, fabricated insights, or recommendations built on incomplete signals—but the workflows around them are inherently different. Generative AI typically sits inside a process with regular human review, whereas agentic AI often makes multiple sequential decisions before a human is involved. When an agent acts on flawed data, that single error can compound across each decision—a budget shift, a paused campaign, or a reallocated bid—all before anyone notices. Which is why data readiness, rather than enthusiasm, is what separates teams that benefit from agentic AI from those it puts at risk.
For all the focus on AI (whether generative or agentic), most organizations are still lagging behind when it comes to data readiness. A patchwork of state-level privacy regulations and increasing signal loss have already made first-party data critical, and AI adoption further raises the stakes.
Clean, consented, and unified data is what ultimately powers accurate insights and effective personalization. Yet few organizations are currently treating it that way: Just 21.4% of industry professionals say first-party data is foundational to their AI efforts, while more than a third admit it plays little or no role.
That gap has proven to be a significant impediment to agentic AI adoption, with eight in ten companies currently citing data limitations as a barrier to scaling agentic AI. And this challenge only grows as advertisers add more tools to their tech stacks, as each new source adds yet another potential point of fragmentation.
Even when teams recognize the importance of first-party data, the data they have often isn’t ready to deliver. Roughly 34% of industry professionals say their first-party data is limited and fragmented, and less than one in five describe it as extensive and well-structured.
Some of the key challenges that hinder marketers’ ability to effectively leverage data are data accuracy and quality, as well as scale and volume of data. While most teams have access to large amounts of data, fragmentation prevents them from forming a reliable single view of the customer. For example, a travel company might struggle to connect loyalty profiles with search behavior and purchase history. Despite having plenty of data, it’s difficult to use and more prone to errors because it’s siloed across platforms and systems. AI trained on only one slice of that picture could misread intent, recommending irrelevant offers or overvaluing the wrong audiences. Multiply that problem across numerous systems and campaigns, and marketers lose both efficiency and accuracy to data problems.
AI delivers its strongest results when it runs on a solid data foundation. Advertisers that prioritize clean, consolidated, and accessible data systems see stronger AI-powered targeting, personalization, and optimizations. For agencies, reliable data ecosystems fuel creative and strategic outputs that capture the nuances of their clients’ audiences.
This groundwork starts with first-party data itself: ensuring it is collected with consent, stored securely, and structured in ways that make it easy to analyze and share across teams. Data hygiene practices, such as regularly auditing for accuracy, de-duplicating records, and unifying customer identifiers, are also essential for maintaining quality at scale. Leaders that embed these practices into ongoing workflows, rather than treating them as one-off clean-up projects, see compounding benefits over time.
Equally important is making data actionable. Tools that unify disparate data sources and streamline reporting consolidate scattered signals into one source of truth, giving AI tools consistent inputs and reducing the errors that come from siloed systems. This also creates a shared foundation across marketing, sales, and finance, making it easier to align strategy and measure impact.
The stakes climb higher with agentic AI. Some platforms now build a unified data foundation directly into the system, so that agentic capabilities draw on consolidated, accurate inputs from the start. Even then, the inputs are only as good as the underlying data. Those foundations matter more, not less, as execution becomes more autonomous.
AI has quickly become embedded in marketing workflows, but its value depends largely on the data that fuels it. Too often, that foundation is fragmented or incomplete.
Organizations that invest in systems to ensure clean, compliant, and unified first-party data will be positioned to capture AI’s full value. The payoffs are significant: stronger ROI, personalization that resonates, and long-term differentiation. In the years ahead, as AI shifts from assisting to acting, those who have built the strongest data foundations are likely to come out ahead.
__
Looking for even more insights into the state of AI in marketing? We surveyed marketing and advertising professionals from leading agencies and brands for our third annual AI and the Future of Marketing report. It’s filled with insights to help industry leaders evaluate how to use AI responsibly, strategically, and with urgency.