E-commerce giant Wayfair goes all-in on generative AI, partners with Google, OpenAI and more

Wayfair, an American e-commerce company based in Boston, Massachusetts, that sells furniture and housewares online, is undergoing a bit of a generative AI makeover. There’s no way around it.

The company recently partnered with Google Cloud and is working closely with them to optimize its operations on the platform. “What’s great about Google is their focus on AI, ML, and generative AI,” said Fiona Tan, CTO at Wayfair, in an exclusive interview with OBJECTIVE.

“We can use Google models like Gemini, but also third-party models like Claude, which can give better results than others in certain cases.”

Interestingly, Wayfair also has a separate partnership with OpenAI.

The company’s move to a fully cloud-based model means it’s ready to adopt new models as they evolve. “If something isn’t working today, we can try again in a few weeks and it could be much better,” she noted.

Wayfair currently has over 22 million active customers and offers over 30 million products from over 20,000 suppliers in the U.S., Canada, Germany, Ireland and the UK.

What about India? Rohit Kaila, chief technology officer at Wayfair India, sees an opportunity for AI, especially generative AI, to bridge the gap in a country with so much diversity.

“This is a great opportunity for us to leverage AI, especially generative AI, to adapt to local nuances,” Kaila said.

If we think about handling something as large as Flipkart’s catalog across an entire organization, the AI-based tools available today are far superior, enabling much faster turnaround times and significantly improving the user experience.

“Previously, companies would group customers into five or six categories and tailor experiences to each one. Now, generative AI is enabling Wayfair and other companies to create personal shopping journeys for each customer—every time they click,” he said.

Pioneering AI-powered search

Wayfair says it has always been at the forefront of using AI and machine learning to transform e-commerce search to help customers find and connect with products. “In e-commerce, AI and ML are widely integrated across several areas. Search functionality is key, helping to understand customer intent and match products to search queries,” Tan said.

Beyond search, AI is revolutionizing performance marketing, where it precisely matches spend and ensures ads reach the right audience.

Tan believes the potential has only gotten bigger as generative AI enters the industry. She said AI can delve deeply into customer sentiment and use that insight to refine product offerings based on massive amounts of data.

Wayfair has been using AI and machine learning for a long time, particularly for decision-making, forecasting and optimization.

According to Tan, Wayfair was the first to apply these technologies to marketing. “Given how much we invest in customer acquisition, it was critical to ensure that we were optimizing our spend. We started by building models that helped us identify the channels that were driving the most traffic, maximizing the return on our advertising investments.”

In addition to marketing efforts, the challenge is also helping customers navigate the vast array of non-branded products.

Tan believes that shopping for household goods presents unique challenges compared to more direct purchases like faucets. It’s often difficult for customers to specify exactly what they’re looking for, which makes it difficult for e-commerce platforms to provide accurate search results. Wayfair also faces additional hurdles due to the often incomplete or limited data that vendors provide about their products.

For example, when picking up a chair, key information like dimensions, adjustability, or back support may not be detailed enough or may not be specific enough. This is where AI comes in to enhance the search experience by identifying and extracting key product attributes.

Wayfair is hiring advanced multimodal modelsthat integrate both text and image data to solve this problem. Initially developing many of these models in-house, the company has expanded its approach to use generative AI and LLM trained on extensive web data. These models are particularly effective at interpreting product features from images and descriptions, enabling a more nuanced understanding of each product.

By combining visual and text data, Wayfair can leverage new AI capabilities to make product matches and recommendations, offering customers a more tailored and effective search experience.

Wayfair vs. the world

Wayfair is not alone. Its competitors are also focusing on generative AI, in exclusive interaction with OBJECTIVEAmit Kapur, head of AI and Analytics at Lowe’s India, said his company has deployed over 40-50 AI models in production and was an early adopter of generative AI, working with OpenAI even before ChatGPt.

Wayfair isn’t alone in using AI to transform e-commerce. Lowe’s, a Fortune 50 home improvement retailer, has deployed more than 40-50 AI models on its platform to improve customer experiences and operations, as well as build an internal omnichannel order management system.

Few people know that Lowe’s was one of the first to partner with OpenAI, even before the launch of ChatGPT in November 2022, when the data and AI team successfully tuned the GPT-3.5 model to improve the accuracy of product data, reduce friction in online searches, and increase the error detection rate by up to 60%, resulting in a smoother and more efficient shopping experience.

In 2022, IKEA introduced IKEA Kreativ, an AI-powered tool that lets customers design and visualize their living spaces in immersive 3D from any device. Combining IKEA’s interior design expertise with advanced spatial computing and machine learning, it’s a breakthrough in personalized interior design.

Even Target and Walmart have adapted AI to provide customers with a better shopping experience.

E-commerce platforms have turned to artificial intelligence (AI) to navigate the dynamic landscape of AI, real-world applications, and more.