April 24, 2026
If there’s one recurring word that comes to mind when I think about how marketing has evolved in recent years, it’s this: overloaded.
Consumers are overloaded by the messages coming at them from televisions, phones, wearables and billboards. And brands are overloaded by the ever-multiplying ways to reach consumers. It’s become so complex that it’s nearly impossible to put together a cohesive, integrated marketing plan.
AI can help. But only if it’s built the right way.
That’s why it’s taken nearly a year for our team, with the help of our partners, to set up the infrastructure that enables an AI-enabled marketing mix model. Because we know that the output of AI can look impressive. But if you don’t put the time and resources into the back-end capabilities, none of it works.
When our systems fully launch in May, we’ll have the ability to make fast, informed decisions about where our advertising dollars go based on what’s actually happening right now in the market, not what happened months ago.
In the past, we’ve focused on reach.
Now, we’re prioritizing relevance, boosting the efficiency of every dollar spent.
Owning the data, not renting it
Historically, advertising metrics were funneled to us by sellers using their own frameworks, which meant we never had a full picture.
I like to say we were renting our data. Now, we not only own the house, we poured the foundation of the house we built.
We’ve designed internal capabilities to ingest and analyze hundreds of thousands of marketing inputs, giving us full visibility across our ecosystem. That all starts with our partnership with a collaborative analytics platform that acts as our “data washing machine,” continuously pulling in and standardizing data from social platforms, search engines, streaming services and more. What used to take five months of manual sorting in dizzying spreadsheets now happens dynamically overnight.
The data then feeds into agentic-AI powered platform. This is where it becomes meaningful, giving us visibility into all of our sales data so I can see the entire business in real time and ask questions. That’s how I start every day, inquiring about things like: what are my best and worst performing channels? Where should I move my dollars? What’s driving revenue? It gives me the answers immediately and takes in the context of the current macroenvironment. If the first layer is our “data washing machine” this process presses and folds the laundry, selects an outfit and most importantly, recommends what to pivot to when the forecast just changed from snowstorm to refreshing breeze.
Turning data into decisions
While we’ve been building this infrastructure, our marketing strategy has transformed. In the past, we ran our media mix model three times a year. By the time we got the results, we were already planning far into the future. Now, we can read performance every month, using both four- and 12-week lookbacks. That will allow us to see demand capturing, brand-building and demand generation all at the same time. And because it’s been trained on three years of Hershey data across trade and marketing, it knows what signals to pay attention to.
In addition, we used to focus on measuring five brands a year. Moving forward, we’ll be measuring all brands with active media in the entire portfolio. That allows us to buy and optimize all forms of media faster, unlocking growth.
One of the best examples of this shift is an instance we saw with Kit Kat on social media. Historically, we focused on paid media because that’s we saw the biggest return on our investment. Now, we have visibility into both paid and organic, and the difference is clear. Our social team is posting organically to Kit Kat channels every day for an entire year, without taking a break. If you’ve ever managed a social media channel for any brand, much less one of the world’s beloved candy brands, you know what a feat this is. The results so far surprised us. As you’d expect, paid delivered more reach. But platform-native, organic content delivered a view rate that was 4.5 times higher, with 13 times the engagement. This is changing how we think about investing in media (and making our social media team question their bold ideas. You’ve got this team. Only 254 days to go!). In the future, this type of granular data will be available to us immediately, so the team can decide where to invest time and dollars.
In our overloaded world, reach is just a number. Relevance is the future. By moving from simple spreadsheets to AI-powered systems, we’re not just working faster, we’re also making better decisions. And more importantly, we’re creating space for more critical thinking—the critical human element that cannot be replaced.
AI will continue accelerating the pace of change in marketing, there’s no doubt. But at Hershey, we’re moving forward with confidence. Because we know we’ve built a solid foundation, and we’re ready for whatever comes next.