At Advertising Week 2025, Top Analytics Leaders Champion Sustainable Data Quality Over AI Hype

At last year’s Advertising Week, the mood was abuzz with excitement and possibility as everyone rushed to imagine what AI could do. This year, it seems the industry has returned to reality. As conversations shift from speculation to substance, most organizations are moving away from lofty aspirations and toward the harsh truths about the data and infrastructure required for AI.

As Manager of Advanced Analytics at Walmart Connect, Aarushi Khera is responsible for translating data into strategy for one of the world's largest companies today. With a career formed at the intersection of data and strategy for major brands like CVS Health and Workday, Khera has a proven track record of leading successful teams and building the AI-driven tools that transform complex data into business impact. She says the conversations on the Advertising Week 2025 stage echoed the same challenges she already faces every day.

For Khera, the next phase of AI adoption is about moving beyond performance and into proof, starting with the admission that AI’s most significant problems are not new. "All the issues that come with AI are the same ones that existed before. The world seemed to forget that AI needs those same fundamental data rules. People thought it would fix everything, but it doesn't," Khera says. AI is only as good as the data it’s trained on, she explains—a fundamental rule that got lost in the hype.

The real test begins when moving from a pilot to scale, she continues. This is the critical turning point that distinguishes "AI theater" from real-world applications, making data governance a practical necessity.

• Expectation vs. reality: Referencing panels like Less A, More I, which focused on activating real data systems, she notes this is where hard questions emerge. "We need a way to understand what AI theater is and what a real AI application is. Some companies want to build their own tools, while others prefer to buy. But there’s always this sense of urgency. You have to make the right decision, and you have to make it fast."

In many ways, the pivot back to fundamentals also reframes the fear of job replacement, Khera says. Because even the most advanced systems cannot originate human intent, they can only amplify it.

• Human in, machine out: Using AI to execute a plan frees people to focus on strategy and insight, Khera explains. But the technology itself is inert, requiring human intellect to give it direction and purpose. "Everything created with AI starts in someone's mind, with a person trying to solve a problem. That's why you need someone to drive the AI tool, because it won't drive itself. At the end of the day, you still need people to build it and guide it."
• Powered by people: To reinforce her point, Khera deliberately looked beyond the AI-focused sessions at Adweek for proof, finding it in stories grounded in timeless human strategy. She points to the Don't Do NYC Dirty campaign and Kolkata Chai Co.'s Brewing Success panel as evidence that fundamentals like authenticity and customer understanding remain the bedrock of success. "Though AI is great, it's just like any other automation tool. Technology on its own doesn’t create impact. It only works when it’s guided by real, data-driven insight."

But reaching that integrated future comes with hard lessons. Instead of a setback, Khera views the high failure rate of AI projects as a necessary learning phase for an industry that has been searching for a quick fix. For her, the failures prove that "there is no magic potion to make things work faster." Ultimately, she envisions a future where a balance between technology and timeless fundamentals defines success. "At the end, you need to make your data foundation really strong."

"I think the conversation will become more integrated," she predicts. "A team might present a new media mix where 10% of the result comes from AI, but the other 90% is still the foundational data and strategy work that has always mattered. Over time, AI will settle into that rhythm and become just another tool we use every day, the way data is now."