OpenAI's quiet integration of native shopping into ChatGPT, coupled with a significant partnership with Walmart, marks far more than just another AI feature rollout. It signals a fundamental shift in how consumers discover, compare, and purchase products online. For the first time, shoppers can browse and buy directly within an AI conversation, bypassing traditional search results, endless scrolling, and marketplace middlemen.
To understand the profound implications for the future of search, marketplaces, and digital marketing, we spoke with Tim Vanderhook, CEO of Viant Technology. Vanderhook, who recently shared his perspective on LinkedIn, believes this move could redefine the entire digital commerce ecosystem, dismantling the "gatekeeper dynamic" that platforms like Amazon and Google have long relied on. He explains why large language model (LLM)-powered shopping could reshape the sales funnel, rewrite attribution rules, and usher in a new era of AI-native marketplaces.
The Dawn of a New Marketplace
Vanderhook characterizes this development as "the beginning of an exciting new kind of marketplace." He sees LLM-powered commerce as a foundational shift, not merely in how people discover, but in how they interact with products, services, and brands. Traditionally, platforms such as Google, Amazon, or Walmart have acted as digital commerce gatekeepers, where visibility is dictated by rankings, algorithms, or marketplace dynamics. In an LLM-powered future, the interface becomes conversational, personalized, and significantly more dynamic.
This model re-centers discovery around intent, moving beyond mere keywords. Instead of generic search results, consumers will engage with AI-driven shopping assistants that understand context, including the "where, when, why, and for whom" of their purchase. This effectively collapses the traditional "search – click – checkout" funnel into a single, intelligent conversation.
For marketers, this means success will hinge on the quality of engagement and product fit, rather than solely on ad spend or search rankings. In many ways, it's the inverse of the search economy: instead of bidding for prime space, brands will need to earn their relevance through compelling storytelling, robust brand-building, and fostering trust.
Breaking Down the Gatekeepers
Vanderhook asserts that OpenAI's move could "break down the gatekeeper dynamic" that Amazon, Walmart, and others have long relied on, signaling a real power shift in digital commerce. He confirms that this shift is already underway. Legacy players like Amazon have historically benefited from their control over both inventory and discovery. This changes dramatically when the primary discovery interface moves from their proprietary search bars to independent, intelligent LLMs like ChatGPT.
However, he cautions against counting out the incumbents entirely. These companies have built massive infrastructure and consumer trust. Many will adapt quickly by integrating with LLMs or embedding their services into new ecosystems. Yet, the power dynamic will undeniably shift: from owning the entire sales funnel to participating in a more open, orchestrated marketplace. In this new environment, the advantage will go to whoever can deliver the best outcome for the consumer, not just whoever owns the virtual shelf space.
The Evolving Role of Brands and Marketers
If LLMs become the primary interface for discovery and transactions, brands and marketers face a seismic change. When product discovery becomes conversational and personalized—no longer driven by static rankings or paid placements—traditional media strategies require a new playbook. Brands must optimize not just for keywords, but for context. This elevates the importance of full-funnel advertising, tailoring paid media strategies around intent, and ensuring retail media campaigns can be activated, optimized, and measured in real time.
In an LLM-driven world, one of the most reliable ways to guarantee visibility is for consumers to ask for a brand by name. Vanderhook highlights that most marketers still allocate nearly 70% of their paid ad budgets to "Demand Capture" channels like search and social, which harvest existing intent. Only 30% is spent on long-term "Demand Generation" and new business growth through channels like Connected TV and streaming audio. This ratio made sense in a keyword-driven world, but in an AI-driven one, marketers gain the power to shape the very conversations that define their brands.
The brands consumers already know and trust are the ones most likely to appear in an LLM's response. Companies that succeed in the LLM era will flip this script, investing more in brand building, Connected TV, and storytelling—the work that generates demand before a consumer ever types (or prompts) a query. In this new landscape, brand storytelling becomes a crucial visibility strategy.
Partnerships Now, Disintermediation Later
Vanderhook suggests that in the short term, marketplaces will partner with OpenAI, but in the long term, OpenAI may not need them. He explains that in the short term, the relationship is symbiotic: marketplaces provide supply, fulfillment, and customer trust—elements LLMs need for the "last mile" of commerce—while OpenAI provides access to intent at scale. Both sides benefit.
However, long-term, LLMs could evolve to connect directly with retailers, effectively cutting out the middle layers. This opens the door to new business models, such as "preferred placement" fees within conversations, affiliate commissions, or verified product data partnerships. Smaller retailers, in particular, stand to benefit significantly. Historically, they have struggled to compete on the first page of Amazon or Google. In a conversational model, they can plug into the system via APIs and win based on merit, product value, or relevance, rather than simply outspending competitors on advertising.
The Future of Attribution and Advertising
In an AI-native commerce model, the traditional sales funnel collapses. Search and purchase often occur in the same moment, necessitating an evolution in attribution. Brands will require systems capable of measuring the full path from prompt to purchase, across various channels and devices.
In this new world, marketers must move beyond chasing last-click metrics and instead optimize for true incrementality. The key questions become: "What initially drove the purchase intent?" and "How can we replicate that upstream influence?" This focus on understanding and influencing the entire customer journey represents the future of attribution.
Trust, Transparency, and Brand Safety
If ChatGPT becomes a transactional interface, issues like brand safety, product authenticity, and trust become paramount. Vanderhook believes consumers will rely on AI-driven recommendations, but "if and only if, the system earns that trust." This makes brand safety, transparency, and authenticated data non-negotiable.
LLMs will need robust accountability controls, clearly indicating the product's origin, how it was vetted, and its authenticity. They will also need to demonstrate their reasoning—not just "what" they recommend, but "why." Consumers are already skeptical of black-box recommendations, so AI must be explainable and accountable.
For brands, this means actively owning their presence within the AI ecosystem. They must provide structured data, ensure their offers and inventory are verifiable, and align with partners who prioritize identity, measurement, and integrity. As AI reshapes the interface of commerce, these values will only become more essential.
What Marketers Should Do Next
As Vanderhook emphasizes, the rise of LLM-driven shopping isn't just about introducing another channel; it fundamentally redefines how intent, discovery, and conversion intersect. For marketers, this means preparing for a world where visibility depends less on search rankings or ad placements and more on how effectively their data, product information, and brand trust are integrated into AI ecosystems.
The winners in this new landscape won't be those who merely chase algorithms, but those who make their brands intelligible—and indispensable—to intelligent systems.
