Join top executives in San Francisco July 11-12 and learn how business leaders are getting ahead of the generative AI revolution. Learn more
Online furniture store wayfair embraces the power of generative AI with a thoughtful, measured approach involving a stakeholder council and core thesis to help accelerate business productivity.
In a session at today’s VentureBeat Transform 2023, Wilko Schulz-Mahlendorf, head of pricing and marketing science at Wayfair, provided insight into how the company is using generative AI today and its strategy for integrating more AI in the future.
>>Follow all of our VentureBeat Transform 2023 coverage<
A key tenet of the Wayfair approach to generative AI is to take a careful and thoughtful strategic look at the technology before rolling it out to production; it is an approach that also involves people.
“At Wayfair, we have strict quality standards and we don’t want to show our customers or suppliers content that could involve hallucinations,” said Schulz-Mahlendorf. “We really put a bounty on people in the loop.”
Giving Wayfair employees ‘super powers’ with generative AI
As part of Wayfair’s approach to generative AI, the company looked at applications where the technology could boost workforce productivity — what Schulz-Mahlendorf called giving employees “superpowers.”
Schulz-Mahlendorf explained that Wayfair identified some specific tasks where gen AI technology could help its inbound sales and service teams. Those tasks include text summarization, product recommendations, and agent suggestions for best actions.
Wayfair also needs and writes a lot of content for its site, a critical area where generative AI helps.
“We’re limited by how many copywriters we have in terms of how much text we can actually put out,” Schulz-Mahlendorf said. “We’ve been evaluating a number of different products, including some of the vendors who are here at the conference, to see if we can double or triple the efficiency of our human copywriters.”
Generative AI copy generation is not intended to replace humans, but rather to help them be more productive. Schulz-Mahlendorf said the goal is to generate an initial design that can then be polished and refined by humans to meet Wayfair’s exacting standards.
“There’s a lot of empty space on many websites that went unfilled before, where we can now create engaging content for our customers,” said Schulz-Mahlendorf. “It’s not about replacing content that already exists, it’s about putting content in places where we might not have had anything before.”
How Wayfair decides where to use generative AI
Schulz-Mahlendorf said Wayfair has assembled an internal generative AI council to help evaluate the technology’s strategy and potential use.
He explained that the board is a group of people drawn from different business areas of the company. It helps evaluate the terms of individual technologies, implementation ideas and business value.
The council also helps determine whether Wayfair should buy or build gen AI technologies as the company develops its strategy. Schulz-Mahlendorf stressed that to date, in terms of vendor technologies, there is no clear winner across the board for all of Wayfair’s use cases.
“We take a very pragmatic approach, which is let’s think about the cost of each of these licenses, let’s think about what each of these things can do best, and let’s work with suppliers and partners who are willing to work with us. working together… to adjust things along the way,” he said.
VentureBeat’s mission is to become a digital city plaza where tech decision makers can learn about transformative business technology and execute transactions. Discover our Briefings.