Hershey Throws AI Agents at Its $2 Billion Marketing Mess — Will It Actually Work?

Hershey, the candy colossus, just admitted it has been flying blind on a massive chunk of its marketing spend — to the tune of $2 billion. The solution? Partnering with Mutinex and Tracer to overhaul their Media Mix Modeling (MMM) using so-called ‘agentic AI.’ Translation: instead of the usual sluggish, spreadsheet-driven guesswork, Hershey is betting on AI agents to speed up decision-making and optimize media budgets in ways that actually move the needle.

Let’s be clear: MMM has been a known bottleneck for years. The traditional models are slow, often outdated by the time they deliver insights, and notorious for their inability to capture complex, real-time market dynamics. Hershey’s scale only magnifies this failure. The company’s move to inject agentic AI into this process isn’t just a tech upgrade; it’s an admission that their existing marketing analytics were basically a half-assed shot in the dark.

But before we crown AI as the marketing messiah, remember this: the industry has been sold on the AI hype train for years. What Hershey is doing could either be a breakthrough or just another case of fancy tools failing to fix lazy data practices. Mutinex and Tracer promise compressed timelines and smarter budget allocation, but the devil’s in the execution. Will these AI agents actually integrate with real-time data feeds and complex attribution models, or will this be another elaborate PR stunt?

The candy giant’s bet also exposes a broader problem: brands still cling to outdated MMM frameworks that don’t reflect today’s fragmented digital landscape. The $2 billion blind spot isn’t just Hershey’s problem; it’s an industry-wide embarrassment. If AI can genuinely fix this, it will force agencies and marketers to rethink their entire approach to media planning. If it can’t, it will be yet another example of AI being used as a buzzword to cover up systemic incompetence.

Here’s the uncomfortable truth the marketing world hates to admit: AI isn’t magic. It’s a tool that requires rigorous data hygiene, clear objectives, and ruthless accountability. Hershey’s experiment should be a wake-up call — stop pouring money into legacy models and lazy agencies that rehash the same tired assumptions. If you want to move the needle, invest in real-time data infrastructure, ditch the bullshit 10x agencies, and demand transparency from your AI vendors.

In short, Hershey is taking a gamble on agentic AI to fix a colossal blind spot in its marketing spend. For the industry, the lesson is clear: either evolve or keep wasting billions on models that don’t work. And no, slapping “AI” on your MMM won’t magically fix that.