AI Is Moving from Managing Data to Making Decisions With It
Mihir Nanavati and analyst Doug Laney on AI agents, first-party data strategy, and why most MarTech stacks aren't built for what's coming.
BlueConic's Mihir Nanavati sits down with Doug Laney, Research & Advisory Fellow at BARC, author of Infonomics, on EM360's Tech Transformed podcast to unpack the shift from data-informed to AI-directed marketing decisions.
For years, the promise of customer data was straightforward: collect more, understand more, act better. The problem was always the gap between understanding and acting. Too slow, too fragmented, too dependent on systems that weren't built for speed.
AI agents are the first real answer to that gap. Not because they automate more tasks, but because they reason.
"Machines can reason. That is fundamentally different," says Mihir Nanavati, GM and Product Executive in MarTech and AdTech, BlueConic.
The Fragmentation problem doesn't go away with AI
Every MarTech and AdTech stack today has the same structural flaw: each tool optimizes locally. Email optimizes for opens. Paid channels optimize for clicks. On-site personalization optimizes for session engagement. None of it is coordinated at the customer level.
Mihir calls it a paradox of choice. The more tools you add, the worse the customer experience gets. And AI amplifies that problem rather than solving it, because AI agents need context to make good decisions, and fragmented stacks don't provide it. As Nanavati emphasizes:
"AI doesn't need perfect data in many cases, but it needs context."
First-party data is fuel for AI decisioning
The shift away from third-party data has mostly been framed as a compliance response. That framing undersells what's actually at stake.
- Declared preferences
- Behavioral signals
- Purchase history
- Real-time identity resolution
These are what give AI agents the context to make decisions that are fast enough and accurate enough to drive revenue. Treat first-party data as a checkbox, and you're building a decisioning engine without fuel.
Winning enterprises aren't running more experiments
They're the ones that learn fastest and scale what works. Nanavati's direction to leaders is simple:
"Learn very rapidly. Then scale what you've learned."
That requires an operating model where customer context is centralized, AI has the right signals to act on, and decisioning happens at the customer level—not the channel level.