The AI Frontier: How Whatnot’s Acquisition of Shaped Signals a Shift in the U.S. Live Commerce Landscape

By PYMNTS | July 17, 2026

The digital storefront is undergoing a metamorphosis. What began as a nostalgic nod to the television home shopping networks of the 1990s has evolved into a sophisticated, high-velocity digital ecosystem. Live shopping—a format where hosts demonstrate products, engage with viewers in real-time, and facilitate instant, one-tap transactions—is effectively collapsing the traditional sales funnel. By compressing the distance between product discovery and the final transaction to mere seconds, the format is redefining consumer expectations.

Yet, despite its massive success in Asian markets, the format has struggled to achieve similar ubiquity in North America. While China saw live shopping generate a staggering $900 billion in sales in 2025—a figure rivaling the entire eCommerce market of the United States—Western consumers remain hesitant. Recent data suggests that 68% of North American shoppers have yet to make a single purchase through social media. To bridge this gap, industry leaders are turning to advanced artificial intelligence to solve the ultimate hurdle: the discovery paradox.

The Discovery Dilemma: Why Western Live Shopping Has Stalled

The fundamental disconnect between the success of live commerce in the East and its tepid adoption in the West is not rooted in content quality or host charisma. It is a technical issue of curation and scale. In a digital environment where thousands of streams run simultaneously, inventory shifts by the minute, and flash auctions conclude in a heartbeat, the "static" recommendation algorithms that power traditional online retailers are woefully inadequate.

A shopper who wanders into a live stream at the wrong moment—after a desired item has sold out or before a relevant product is showcased—is unlikely to convert. Without the ability to surface the right stream to the right user at the exact moment their interest peaks, the platform becomes a chaotic landscape of missed opportunities. To transition from a niche hobbyist community to a mainstream retail powerhouse, platforms must master the art of hyper-personalization at machine speed.

Whatnot, the enthusiast-focused live marketplace currently valued at over $11 billion, has positioned itself at the vanguard of this challenge. By processing more than 500,000 hours of live video and millions of user interactions on a weekly basis, the company has recognized that the next generation of eCommerce will not be won by inventory alone, but by the intelligence that connects that inventory to the consumer.

The Strategic Pivot: Whatnot Acquires Shaped

In a move that underscores the critical importance of machine learning in the modern retail stack, Whatnot announced on Wednesday, July 15, that it has acquired Shaped, a machine learning startup specializing in real-time recommendation and search infrastructure. While the financial terms of the deal remain undisclosed, the strategic implications are clear: Whatnot is doubling down on its commitment to algorithmic precision.

Shaped, founded in 2021 by former Meta recommendation systems expert Tullie Murrell, brings a specialized toolkit to the table. The startup’s core competency lies in building infrastructure that can ingest vast amounts of behavioral data and translate it into actionable, real-time user experiences. Under the terms of the acquisition, Murrell will join Whatnot alongside a dozen engineers and researchers to spearhead a new applied AI research group.

This talent infusion is designed to tackle the specific volatility of the Whatnot platform. Unlike an Amazon or a Walmart, where product catalogs are relatively stable, Whatnot’s inventory is in constant flux, dictated by individual sellers and live broadcast schedules.

Chronology of a Tech-Driven Expansion

The acquisition of Shaped represents the latest milestone in a period of aggressive growth for Whatnot. The company’s trajectory over the past eighteen months illustrates a concerted effort to scale its infrastructure alongside its user base:

• Early 2025: Whatnot secures significant capital, with its valuation climbing past the $11 billion mark, driven by its success in niche enthusiast communities such as trading cards, collectibles, and high-end streetwear.
• Throughout 2025: The platform launches over 35 new product categories, signaling a push into broader consumer goods and lifestyle markets.
• First Half of 2026: Whatnot maintains its momentum, introducing an additional 45 categories and surpassing the milestone of 1 billion total orders.
• July 15, 2026: Whatnot officially acquires Shaped, marking its entry into a new phase of AI-driven platform optimization.
• July 17, 2026: Industry analysts confirm that the integration of Shaped’s technology is already being fast-tracked to enhance the real-time responsiveness of Whatnot’s discovery engine.

Supporting Data: The Rise of Agentic Commerce

The decision to integrate AI into the live shopping experience is supported by a growing body of data regarding consumer behavior. According to the PYMNTS Intelligence report, “The Agentic Commerce Deep Dive: Payment Infrastructure and the Path to Trust for Merchants,” created in collaboration with Visa Acceptance Solutions, the appetite for AI-assisted shopping is surging.

• AI-Driven Research: 48% of online shoppers reported using AI tools to assist in researching their most recent purchase.
• Traffic Trends: AI-referred traffic to U.S. retail sites experienced an explosive 393% year-over-year increase in the first quarter of 2026.
• The Trust Gap: The data indicates that consumers are increasingly looking to AI not just for search, but as a concierge-like guide to navigate the overwhelming volume of digital retail options.

This trend validates the thesis that shoppers are no longer content to browse through static lists. They are seeking curated, intelligent, and immediate shopping experiences. By deploying AI that can "watch" a stream, identify the products being discussed, and match those products to a viewer’s historical preferences and current session intent, Whatnot aims to mirror the efficiency of a high-end concierge service.

Official Perspectives: Building a Responsive Ecosystem

The integration of Shaped’s technology is viewed by Whatnot’s leadership as the key to unlocking the platform’s full potential. Emmanuel Fuentes, Vice President of Data and AI at Whatnot, emphasized the shift toward a more dynamic infrastructure in the wake of the acquisition.

"By combining Shaped’s technology with Whatnot’s existing systems, we can make recommendations faster, more responsive, and more personalized," Fuentes stated. The goal is to move beyond simple collaborative filtering—which suggests items based on what others have bought—to a more sophisticated "intent-based" discovery model.

Shaped’s systems excel at synthesizing disparate data points: customer purchase history, real-time engagement signals (such as chat participation or time-spent-watching), and inventory metadata. For a platform like Whatnot, this means that a user who enters a live stream for a "vintage sneaker" auction will be seamlessly guided toward a "rare apparel" stream the moment their interest begins to shift.

Implications: The Future of Mainstream Live Commerce

The acquisition of Shaped is more than just a talent acquisition; it is a signal that live commerce is graduating from a niche retail format into a core component of the mainstream digital economy.

1. The Death of Static Discovery

As AI becomes more deeply integrated into the shopping experience, the concept of a "static homepage" will likely become obsolete. Platforms will instead provide "living" interfaces that curate live video content based on the user’s micro-interests. This will allow smaller sellers to reach hyper-targeted audiences without needing massive marketing budgets.

2. The Standardization of "Shoppable Moments"

If Whatnot succeeds in reducing the friction of discovery, other retailers will be forced to follow suit. The industry is currently witnessing a race to integrate real-time AI, which will inevitably lead to a higher standard of UX across the board. The ability to "buy now" within a live feed will become a table-stakes feature rather than a competitive differentiator.

3. Economic Impact

If the U.S. market begins to mirror the growth seen in China, the economic implications are profound. With $900 billion in potential market scale, the integration of AI could unlock hundreds of billions of dollars in dormant eCommerce value. This will likely spark a surge in investment, not only in marketplace platforms like Whatnot but also in the underlying infrastructure of payment processing, logistics, and real-time content delivery.

Conclusion: A New Era of Retail

The marriage of live shopping and AI is fundamentally altering the psychology of the consumer. By eliminating the "search-and-scroll" fatigue associated with traditional eCommerce and replacing it with curated, real-time entertainment, platforms like Whatnot are positioning themselves to capture the next generation of digital natives.

The acquisition of Shaped is a decisive step in this evolution. By prioritizing the intelligence layer of the user experience, Whatnot is not merely betting on a trend; it is building the infrastructure for a future where shopping is an immersive, highly personalized, and immediate experience. As the gap between content and commerce continues to shrink, the companies that master the art of the "right stream at the right time" will undoubtedly define the retail landscape for the next decade.