Campaigns That Land With Humans And AI Combine Emotional Equity With Verifiable Proof

Marketers built the modern brand around emotion. Mascots, music, story arcs, decades of equity, and the trust that compounds when a campaign hits the right cultural moment aren't going away. What's changed is that the human consumer is no longer the only audience deciding whether a brand earns the consideration. AI agents are now embedded in research, shopping, comparison, and verification, and they evaluate the same brand on entirely different criteria. Brands now need to be emotionally meaningful to people and factually legible to machines, and the gap between those two requirements is where most marketing organizations are currently exposed.

Chan Suh is Senior Partner and Chief Digital Officer at the strategy and marketing consultancy Prophet. There, he leads the firm's digital practice and recently launched MAIA, an agentic AI platform built for marketers. He's a co-author of The Rise of the AI-Powered Consumer, Prophet's new report based on findings from more than 2,400 consumers across the US, UK, Germany, China, and Singapore on how AI is reshaping purchase behavior, brand expectations, and trust. For CMOs, the report's most consequential findings concern what consumers now want AI to do on their behalf, and what that demands of the brands trying to reach them.

"The challenge for marketers is, how do you convey emotional benefits to an uncaring AI? It may not be possible, so you will have to compete on facts," Suh says. The implication for brands is direct. Though the layer that has carried decades of marketing investment still works on humans, it does not work on the agents those humans are increasingly delegating to.

The dual audience problem

Prophet's research surfaces the shift in consumer expectations clearly. People want AI to do more than answer questions. They want it to intermediate and protect. "People want AI to be their advocate when they are negotiating on high-consideration and low-consideration purchases," Suh explains. "They want AI to be their intermediary among competing brands for whatever they're looking for."

That request, he says, creates a new structural reality for brands. The AI agent sitting between the consumer and the marketplace will compare products on attributes it can verify, surface claims it can substantiate, and discount messaging it cannot validate. Emotional positioning still matters for the moments the human is in the loop, but the agent's filter happens first, and it filters on facts. Suh frames the consequence as a credibility problem. "One of the most popular uses of AI for consumers is verifying claims by brands. If you say things and they're not verifiable, you will have discredited your brand."

The research also captures a sharp decline in consumer optimism about AI between Prophet's 2024 and 2026 findings. Consumers have moved into a more pragmatic, in some cases pessimistic, view of what AI will actually deliver, and they've grown skeptical that brands will use AI ethically on their behalf. The skepticism is most pronounced among Gen Z, contradicting the longstanding marketing assumption that younger generations are uniformly the early adopters of new technology. "Brands have to understand that there are significant generational differences, and maybe in unexpected ways,” he notes. It used to be that younger people would adopt the newer things and then it would wash over the rest as they got older. That’s not happening anymore."

When emotional campaigns still work, and when they do not

One path forward that Suh foresees is marketer leaning harder into campaigns that make no explicit claims at all. He acknowledged that pure entertainment can feel like a safe zone. "Sometimes the way around it is humor," he says. "You don’t have to prove funny things. The Geico Gecko, for example, or Budweiser’s commercials, are super emotionally entertaining, and they don’t actually have claims in them. That might be a refuge for marketers."

The approach remains valid, but Suh cautions that the upside it creates is smaller than marketers assume. The same campaign that builds affinity with the human consumer does little for the agent comparing sourcing methodology or ingredient transparency on the consumer's behalf. "If AI is the audience, it doesn't really care about the Budweiser Clydesdales or some green lizard. I just don't think it's that useful." In his view, the corrective is not to abandon emotional work, but to recognize that emotional work alone is no longer a complete strategy. Every brand now needs a verifiable backbone underneath the campaign, capable of standing up to the questions an AI agent will ask on behalf of a skeptical consumer.

What brands need to build

Suh's practical guidance for CMOs starts with a diagnosis. Most organizations don't have a current read on how AI sees them. "Understand what LLMs are saying about you. Understand what some of the heavy users are saying. There are a lot of heavy users and there are a lot of light users. The traditional bell curve is inverse."

The diagnostic work feeds into a broader discipline that goes beyond traditional AEO. The synthesis an LLM produces about a brand is shaped by Wikipedia, Reddit, social media, reviews, third-party publications, and the brand's own content. Optimizing the website is not enough. The entire public footprint becomes part of the input the agent draws on. Because of this, verification has to operate at a level of granularity that most brands have not historically supported.

Suh illustrates with a recent project involving agentic AI questions about a protein source. "If you’re telling me this is 20 grams of protein, prove it to me. Show me where it’s coming from. Is it vegan? Is it vegetarian? And if it’s derived from milk, what methodology was used?" Such queries, he says, aren't edge cases. They're the kind of questions AI now makes consumers sensitive to, and the brand that can't answer them in structured, verifiable form loses the comparison.

The brands moving fastest on this work are already operating in ways that fit the new model. ASOS layered Virtual Try-On and Style Match onto online shopping to address the emotional friction of self-doubt and isolation, blending utility with empathy. Bank of America's Erica functions as a proactive financial copilot, using sentiment models to adjust tone and escalate to a human specialist when needed, executing on what the bank calls "high tech, high touch." Starbucks built Green Dot Assist to augment baristas rather than replace them, giving staff faster access to recipes, equipment troubleshooting, and operational information so they can focus on the hospitality humans actually deliver. None of those brands are leading with an AI label. Instead, they're using AI to deepen the relationship with the customer, which is the move that registers with both audiences.

Customer support is part of the brand system

The deeper strategic shift Suh advocates for is reframing customer support from a cost center into a brand asset. AI gives marketers the operational capacity to do this at scale for the first time, and the consumer research suggests that capacity is exactly where attention should go. "What does support mean? Support in selecting your products, using your products, and preventing problems from happening with your products. It's proactively asking and maintaining that relationship. People don't mind that it's an AI, as long as it's proactive and comes from a caring place," he asserts. 

That definition is broader than the function most organizations currently invest in. It treats every customer touchpoint, before purchase, during use, and across the lifetime of the relationship, as part of the brand. Suh sees that reframing as a direct challenge to how most companies operate today, and an opportunity to capitalize on. "A lot of brands think of customer service as a necessary evil. Companies we consult with want to get their customer service cost down, and their hope is that people never call them because everything is working fine. We need to rethink that," he says. "The less they talk to you, the less they care about you."