Wayfair (NYSE:W) executives highlighted how the company is using generative AI and a recently completed technology re-platforming effort to improve product discovery, supplier onboarding, marketing creativity, and internal productivity, while keeping a focus on a differentiated on-site shopping experience. The comments came during a discussion hosted by Bernstein U.S. Emerging Internet Analyst Nikhil Devnani with Wayfair Head of Investor Relations Ryan Barney and Chief Technology Officer Fiona Tan.
From predictive machine learning to generative AI
Tan said Wayfair has used data and “more traditional predictive ML” for more than a decade, particularly for style-based search and recommendations and for marketing technology that helps optimize ad spend. She framed generative AI as an extension of that foundation, enabling the company to do prior tasks “better” while also opening up “new things that we can do with the capabilities.”
Re-platforming: moving from bespoke systems to cloud-based building blocks
Asked about Wayfair’s multi-year tech re-platforming initiative, Tan said the company shifted from an “on-prem solution” to the cloud and moved toward “modern decoupled platform architectures.” She said the older stack was built for “entrepreneurship and speed” but resulted in “bespoke experiences” that became difficult to maintain at scale.
The newer approach emphasizes reusable components, reliability, performance, and an “API-driven” foundation that is “agent-ready.” Tan added that the change supports faster testing and iteration, and improves developer productivity by enabling use of cloud-native components.
Supplier onboarding and catalog enrichment with frontier models
Wayfair works with roughly 20,000 suppliers, and Tan said generative AI and multimodal “frontier models” have enabled the company to scale catalog enrichment in ways that were not previously possible. She described using these capabilities to help suppliers onboard faster by enriching, augmenting, correcting, and validating product information.
Tan said Wayfair works “very closely” with both Gemini and OpenAI, and noted that OpenAI recently published a case study on Wayfair’s catalog work. She also said Wayfair has integrated product feeds directly to Google and OpenAI so those platforms can access “the most up-to-date, most detailed, accurate product information, inventory information, location, et cetera.”
In addition to improving on-site shopping, she said more complete and accurate product data can help when AI platforms “are say scraping our site,” and that Wayfair is also collaborating with partners on which “additional attributes” are useful for AI-native discovery experiences.
Customer experience: visualization, Discover, and the path to “Stylist”
Tan said Wayfair has seen “some measurable lift” tied to AI-enriched listings and improved imagery and accuracy, though she did not quantify the impact. She described a shift from supplier-provided product photos to AI-enabled visuals that can show items in room settings and highlight complementary products across Wayfair’s catalog—work that previously required costly photo shoots.
Tan pointed to the company’s “Discover tab” as an example of a new experience that is driving conversion lifts among engaged users. She also discussed an upcoming product—referenced as seen at Shoptalk—that would present room scenes “with real products that are buyable” on Wayfair, and allow customers to upload a photo of their space to visualize products in context.
On the future of on-site search, Tan argued that e-commerce has long relied on a “search bar and some filters,” which is limiting for a visual category. She described an ambition to move beyond longer queries to multi-turn conversational shopping that can incorporate voice, imagery, video, and persistent context across pages. She positioned “Wayfair Stylist” as part of that broader effort to infuse assisted experiences throughout the commerce journey.
Traffic from AI platforms, UCP integration, and profitability implications
When asked about consumer behavior coming from platforms like Gemini or ChatGPT, Tan said the channel is “still early days” and “nascent,” but “growing quickly.” She characterized that traffic as high-intent, saying it “converts well” because users often arrive after doing research and receiving recommendations within those platforms.
Barney declined to quantify the percentage of traffic from those sources, calling it “very small.” He said Wayfair wants to “be where our customers are at,” similar to how the company has participated in Google Search over the past two decades.
Tan said Wayfair was a “founding partner” with Google on UCP, which she described as a broad commerce protocol that started with product discovery and checkout and is expanding into pre-purchase capabilities such as “loyalty or promotions.” She said Wayfair’s modular architecture made it easier to support these protocols. On OpenAI, she said the company is focused primarily on discovery, noting OpenAI’s emphasis has shifted away from “instant checkout” toward helping users find products.
Barney addressed investor concerns about potential erosion of direct traffic, saying Wayfair intends to show up on major AI platforms while keeping a “differentiated and quite unique” on-site experience tailored to home. He also cited Wayfair’s scale—“over 40 million products”—and said AI tools can improve curation and personalization to help customers navigate a large catalog.
On marketing, Barney said generative AI has been used to speed ad creative development, including short-form video, and to enable faster iteration that would not be cost-effective with traditional production. Both Barney and Tan added that Wayfair’s marketing technology remains highly sophisticated, with experimentation as a core capability.
Tan said Wayfair is also using generative AI in customer service, balancing higher autonomous “containment” with customer satisfaction and fast resolution. She emphasized that the company still expects “super powered human” agents to handle complex interactions, which she said are common in the home category.
On internal productivity, Tan described a shift in software engineering from using copilots to “coding agents” that can support broader software-development workflows. She said many engineers are moving from “being in the flow of coding” to “orchestrating,” while product managers and designers also adopt the tools to avoid bottlenecks.
Barney said Wayfair has pursued cost efficiency for years and believes it can grow without expanding headcount, creating leverage in the fixed cost base. He referenced “roughly $12 billion of sales in 2025” and said the team in place can support a larger business. Asked where AI benefits may show up over the next one to three years, Barney said he expects the impact to be “mostly on the top line,” with profitability flow-through aided by keeping fixed costs “more or less static.”
Looking at 2026, Barney said initiatives that helped performance in 2025—including the loyalty program, Wayfair Verified, and influencer-based marketing—are continuing to compound, and he expects additional customer experience improvements driven by generative technology to further differentiate Wayfair and contribute to a widening “share spread” versus the overall category.
About Wayfair (NYSE:W)
Wayfair Inc (NYSE: W) is an e-commerce company focused on home furnishings and décor. Through its platform, Wayfair offers a broad assortment of furniture, lighting, home textiles, kitchenware and decorative accessories. The company’s portfolio includes flagship sites such as Wayfair.com, as well as specialty retail brands like Joss & Main, AllModern, Birch Lane and Perigold, each catering to distinct design styles and price points.
Founded in 2002 by Niraj Shah and Steve Conine under the name CSN Stores, the business rebranded as Wayfair in 2011 and went public in 2014.
