Research & Commentary — Gina Cui, Ph.D.

Commentary Harvard Business Review

How AI Agents Are Breaking the Platform Business Model

Cui, Y., van Esch, P., & Kietzmann, J. · Harvard Business Review · April 13, 2026

What happens when AI agents start doing the browsing, comparing, and buying for us? In our new piece for Harvard Business Review, Jan Kietzmann, Patrick van Esch, and I argue that the consequences for digital platforms are existential — not incremental.

Platforms like Google, Amazon, and Meta built their empires on a simple premise: consumers browse, and while they browse, they see ads. But AI agents don't browse. They execute. They complete purchases from intent to fulfillment without ever loading a feed, clicking a banner, or scrolling past a sponsored listing. That means advertising revenue — 75% of Google's income, 97% of Meta's — faces a structural threat, not a cyclical one.

The data backs it up. During Cyber Week 2025, AI agents influenced $67 billion in global sales. AI traffic surged 805% year-on-year on Black Friday. Nearly half of Gen Z and Millennial shoppers delegated their holiday purchases to an AI agent entirely. This isn't a niche behavior anymore — it's a tipping point. Platforms can resist (legal battles, like Amazon vs. Perplexity), adapt (build their own agents, risking cannibalization), or reinvent as agent-ready infrastructure. But doing nothing is not an option.

Key takeaway: AI agents don't see ads, don't comparison-shop on your platform, and don't stay loyal to your ecosystem. The business models that powered the internet for two decades are being quietly dismantled — and the platforms that survive will be the ones that redesign for an agent-first economy.

Read the full article: Cui, Y., van Esch, P., & Kietzmann, J. (2026, April 13). How AI Is Threatening Platforms' Revenue Streams. Harvard Business Review. Read on HBR →

Commentary The Conversation

The $22 Smoothie Paradox: Why Economic Anxiety Drives Premium Spending

Cui, Y. & van Esch, P. · The Conversation · April 8, 2026

Consumer confidence is at a decade low. Food prices are up nearly 30% since 2019. And yet Erewhon — the Los Angeles grocer selling $22 smoothies — is generating up to five times the revenue per square foot of a typical supermarket. How does that add up?

In a new piece for The Conversation, Patrick van Esch and I argue it adds up precisely because the economy feels out of control. When the macro story is bleak, people reach for small, bounded luxuries they can actually control — a glass bottle of cold-pressed juice, an heirloom tomato, a jar of ghee with a story on the label. Consumer psychologists call this compensatory consumption: buying something to restore a sense of agency that life has taken away. The lipstick index after 9/11 was one version of it; the $22 smoothie is the 2026 version.

What makes the specialty food category distinctive is that each purchase arrives pre-loaded with moral justification — it is healthier, it is sustainable, it is authentic. You aren't indulging; you're investing in yourself. And once the purchase gets photographed and posted, it stops being consumption at all and becomes performance. That's the real paradox: the same $22 buys groceries, identity, and an alibi, all at once — but only if you can afford to spend it in the first place.

Key takeaway: When people feel they can't control the economy, they reach for things they can. Premium food is compensatory consumption wearing a wellness halo — and the inequality of who gets to participate is part of the story, not a footnote to it.

Read the full article: Cui, Y. & van Esch, P. (2026, April 8). Why Americans are buying $22 smoothies despite feeling terrible about the economy. The Conversation. Read on The Conversation →

AI & Consumer Behavior

The End of Search as We Know It

Cui, Y. & van Esch, P. (2025) · Business Horizons

What happens when AI agents do your shopping for you? We explored how the rise of AI-powered search and shopping assistants is fundamentally changing the way consumers discover products and make purchasing decisions.

In what we call the "delegate economy," consumers increasingly hand off decisions to AI agents. But here's the catch: only brands that AI systems recognize and recommend will survive. If an AI shopping assistant doesn't know your brand exists, neither will the consumer.

Key takeaway: Brands need to start optimizing not just for human consumers, but for the AI agents that increasingly mediate consumer choices. The rules of discovery are being rewritten.

Cui, Y. & van Esch, P. (2025). The End of Search as We Know It: Only Those Chosen by AI Agents Will Survive in the Delegate Economy. Business Horizons.

Pro-Social Behavior A* Journal

Why Giving to Distant Strangers Makes You Happier

Das, G., van Esch, P., Jain, S. P., & Cui, Y. (2023) · International Journal of Research in Marketing

Does it matter whether your donation helps someone in your neighborhood or someone across the world? We found that it does — but not in the way you might think.

Our research revealed that donors experience greater happiness when giving to socially distant beneficiaries (people far removed from their own social circle) compared to socially close ones. The key mechanism? Benevolence. Giving to distant strangers activates a stronger sense of selfless generosity, which amplifies the "warm glow" of giving.

Key takeaway: Charities can boost donor satisfaction by emphasizing the social distance of beneficiaries. Framing donations as helping distant communities — rather than local ones — can paradoxically increase both the willingness to give and the happiness derived from it.

Das, G., van Esch, P., Jain, S. P., & Cui, Y. (2023). Donor Happiness Comes from Afar: The Role of Donation Beneficiary Social Distance and Benevolence. International Journal of Research in Marketing.

AI & Financial Decisions Sole-Authored

Would You Trust a Robot With Your Money?

Cui, Y. (2022) · International Journal of Bank Marketing

We're increasingly getting financial advice from AI assistants. But does making an AI look and feel more human-like change the financial risks we're willing to take?

My research found that when AI financial advisors are anthropomorphized — given human-like names, voices, or appearances — consumers become more risk-averse. In other words, a more "human" AI makes you play it safer with your money. This challenges the assumption that humanizing AI always makes consumers more comfortable and trusting.

Key takeaway: Making AI look more human doesn't always increase trust. In financial contexts, it can actually make people more cautious — which has important implications for how fintech companies design their AI interfaces.

Cui, Y. (2022). Sophia Sophia Tell Me More, Which is the Risk Free Option of All? AI Anthropomorphism and Risk Aversion in Financial Decision-Making. International Journal of Bank Marketing.

Political Psychology A* Journal

Your Politics Shape How You Feel About AI

van Esch, P., Cui, Y., Das, G., Jain, S. P., & Wirtz, J. (2022) · Annals of Tourism Research

Not everyone feels the same way about AI-powered services. We discovered that your political ideology — whether you lean liberal or conservative — significantly predicts how you respond to AI in service settings like tourism.

Conservatives, who tend to value tradition and personal control, showed more resistance to AI-driven services. Liberals, more open to novel experiences, were more receptive. This pattern held across multiple studies and contexts, from AI hotel concierges to automated travel planning.

Key takeaway: One-size-fits-all AI deployment doesn't work. Companies should consider the political and cultural leanings of their customer base when deciding how prominently to feature AI in the customer experience.

van Esch, P., Cui, Y., Das, G., Jain, S. P., & Wirtz, J. (2022). Tourists and AI: A Political Ideology Perspective. Annals of Tourism Research.