Wayfair CTO to Bring AI Into the Virtual Living Room

Optimization plays an important role in business development, whether by adopting new technologies, rethinking operational workflows or introducing new products and services.

Among all the technologies available today, businesses looking to unlock greater optimization capabilities are increasingly turning to artificial intelligence (AI). In retail and eCommerce alone, AI and generative AI applications are changing the way consumers shop, businesses operate and supply chains function.

“The speed of development of AI capabilities, from text models to multimodal and reasoning agents, is remarkable,” Fiona TanCTO of Wayfairsaid during a discussion for the PYMNTS series “When Chatbots Go Shopping: How GenAI Is Shaking Up the Retail Status Quo.”

The AI ​​revolution is just beginning. The potential for continued improvements is huge, as generative AI continues to evolve and integrate deeper into core eCommerce operations.

At Wayfair, AI is not just a tool; it is part of the home appliances and fabric furniture platform. Tan describes how AI is being integrated into teams rather than being kept in a separate function. “Our machine learning scientists work with engineers and product managers to solve customer problems and operational challenges,” he said. “AI is just one of the tools we use to solve these issues.”

Access to Generative AI is further democratizing its use across the organization. “It’s no longer limited to the tech team,” Tan said. “To fully utilize generative AI, we need to focus on managing change across the organization. It’s not just about the technology but also about enabling people to use it effectively.”

How AI is Changing Online Shopping

Wayfair’s integration with AI is not new. Tan said the company began exploring traditional AI and predictive models more than a decade ago to address complex eCommerce challenges.

“One of the first applications is around our sales channels,” Tan said. “We use predictive models to autonomously adjust our marketing spend and improve attribution accuracy.”

The company’s focus on AI is, in part, driven by the unique challenges of selling home furnishings online. These early investments laid the foundation for Wayfair’s advanced use of AI today.

“The home category is emotional, style-based, and lacks the benefit of well-known brands,” Tan said. “Combined with supply chain complexities and the difficulty of returns, we need to push the envelope with AI and machine learning to create effective solutions.”

With the advent of generative AI, Wayfair increased its capabilities, especially in customer-facing applications. One example is the Deify feature, which uses generative AI to inspire customers with hyper-personalized visualizations. “Customers can upload a photo of their room and see it styled in different ways,” Tan said. “We use text-to-image generation to create these visuals and then match them with real-world products in our catalog.”

“The conversion rate from customers using this feature is very high,” he added. “It bridges the gap between the created world and the real world, making shopping a stimulating and seamless experience.”

Generative AI Takes Stage of the Future of Retail

AI also strengthens Wayfair’s relationship with its suppliers. Through predictive insights and generative content, the company helps suppliers showcase their products more effectively.

“We always share insights like predictive winners with our suppliers,” said Tan. “Now, with generative AI, we can create more personalized content that represents their products better, promoting a win-win partnership.”

Despite AI’s transformative power, Tan emphasized the importance of keeping a human in the loop for critical applications. “We always ensure human management in areas such as customer service or catalog management,” he said. This approach balances innovation with quality control, ensuring that AI-generated outputs meet Wayfair’s high standards.

“We’ve learned that while custom models are better in some areas, general models can outperform them, depending on the use case,” he said.

However, Tan noted that in some areas – such as generative imagery for customer inspiration – a hands-off approach can work. “Customers are surprisingly forgiving of minor flaws, like small table legs, because the overall benefit outweighs the imperfections,” he says. “It’s about finding the right balance.”

As for the future, Tan sees endless possibilities. “We’re excited to lean more into multimodal capabilities and explore how generative AI can inspire customers in new ways,” he said. “There are exciting times ahead, and we’re focused on what’s possible.”