Google Changemakers: Bunker Was Part of the Conversation Shaping the Future of AI in Marketing

The event

Three days, one question: who builds what comes next?

The agenda covered Next-Gen AI, Measurement, Creative Storytelling with AI and The New Economics of Advertising. Four distinct areas that, for Martin, point to the same question: the value model of marketing is changing, and the advantage goes to whoever has the tools and data to measure that change.

"Very intense days alongside leaders from agencies and technology partners of Google across the region, debating where our industry is heading." That is how Martin sums it up.

What he brought back from New York

Four ideas that shape where the industry is heading

1. Business models must evolve at the pace of technology. Agencies and partners should focus on models that quantify value before the effort invested.
2. Creativity remains one of the levers that most drives ROI. In the AI era, creativity is the new way to segment.
3. First-party data remains the foundation of everything. It means strengthening data signals and making each brand's value proposition visible, to better feed algorithms, AI tools and the agentic world ahead.
4. Measuring well means measuring real impact. MMM, attribution models and experiments are the tools to understand marketing's contribution and incrementality to the business.

Next-Gen AI

What's coming isn't a tool, it's a way of deciding

The AI conversation was less about which model to use and more about what happens when systems execute part of the buying, targeting and optimization. When the machine decides for you, the question stops being "do I have AI?" and becomes "do I have first-party data, measurement rules and judgment to review what it decides?"

This is where the agentic world Martin points to as the horizon comes in. Agents can already operate parts of the buying cycle. What they don't delegate is interpretation, and that is where the advantage lies.

• The advantage isn't adopting AI first, but turning it into a measurable process.
• An agent is only as good as the signal it receives. Without your own first-party data, it optimizes on what others give it.
• Without incrementality measurement, automation scales what you already see, not what truly contributes.

Creative Storytelling with AI

Producing variants is easy. Choosing which to scale is not

With generative AI, the cost of producing creative assets falls to nearly zero. That opens the temptation to scale volume without judgment. Martin puts the focus elsewhere: creativity is now the new way to segment. Whoever understands which story connects with which audience reaches results sooner.

Creativity remains one of the levers that most drives ROI.

Scale without judgment multiplies noise. Judgment without scale arrives too late. Balancing the two was one of the topics that came up most in the creativity panels, and Martin's conclusion is direct: mastering creative intelligence is what brings brands closer to their results.

The New Economics of Advertising

When AI produces the asset and the algorithm buys, value shifts

If the algorithm buys the traffic and AI produces the ad, value shifts toward two places: proprietary data signals and proof of real impact. Martin distills it into the idea he brought back most clearly: agencies and partners need to focus on models that quantify value before the effort invested.

First-party data

The signal that matters most is the one you control

When algorithms decide who sees what and at what cost, the only thing you can influence is what you feed them: which data goes in, with what quality and with what permission. That is why first-party data has stopped being a compliance topic and become a performance decision.

The question is no longer whether AI will transform the marketing business. It is who will build the solutions that define how.

Martin proposes two paths to put this into practice: strengthening the data signals you already have, and building and making visible the brand's value proposition, so algorithms, AI tools and agents have something valuable to read. That is the starting point of the vision he took to New York: combining proprietary technology with the MarTech ecosystem so data delivers business value, not just reporting value.

Measurement

Measuring well means measuring real impact

With more automation, the risk is confusing correlation with impact. An attribution model tells you which touchpoint accompanied the conversion. An experiment or an MMM tells you how much of that conversion would not have happened without the investment. Martin insists the three tools are needed together: MMM, attribution models and experiments.

That's why at Bunker we combine MMM, attribution and experiments within a single framework. None is enough on its own, and together they give you an answer you can defend in front of your board.

Conclusion

Organize the data, orchestrate the impact

Amid the transformation AI brings to marketing, organizing data, focusing on solutions that deliver value and orchestrating products and services that create real impact on the business matter more than ever.

— Martin Carniglia, Director of Marketing Science at Bunker