Can AI Save eBay?

EBay was one of the fledgling internet companies at the center of the dot-com boom of the mid-to-late 1990s, and the online marketplace has outlasted more than its fair share of competitors who have not been so fortunate (hello AOL, Yahoo). However, eBay has had a tough time of it lately.

EBay’s stock has dropped by more than half over the past year — from an all-time high of $66.29 per share on July 17, 2015, to close at $30.79 per share on Tuesday (Aug. 2). And the company faces stiff competition, not just from massive global rivals, like Amazon and Alibaba, who more and more seem like they want to sell everything and anything under the sun online, but also smaller merchants using eCommerce upstarts and on-demand ordering apps that make it cheaper and easier than ever before to buy an array of products online.

But eBay believes it has identified an area of growth that is the “future of commerce” and an area where it can outpace and outshine its would-be competitors: the development of artificial intelligence.

“We are planting the seeds right now to ensure that eBay is not only relevant but a leader — a disruptor — in artificial intelligence,” eBay CEO Devin Wenig told CNBC in an interview on July 21, the day the company reported surprisingly higher-than-expected quarterly earnings. “Some of the acquisitions we’ve done are world-class teams of both data analysts and engineering capability that are going to allow us to use this incredible data set that we have.”

EBay has been quietly throwing more and more of its weight behind AI development and implementation, and so far, the efforts seem to be paying off. On July 21, eBay reported quarterly earnings of $0.43 per share on sales of $2.23 billion, a jump from what analysts forecasted for the quarter at $0.42 per share on $2.17 billion in sales. EBay’s stock jumped about 11 percent that day.

Wenig has hailed artificial intelligence as the “future of commerce,” and its practical applications to eBay’s massive and sophisticated online marketplace and the troves of data it has collected on customers through the years are pretty exciting to think about.

EBay hopes to use AI to build a “personalized shopping experience” for each one of its users based on browsing data and past shopping history, a sort of concierge that knows that, if you bought or showed interest in this item, then you might buy or be interested in that item.

And eBay has quietly begun assembling the building blocks to do just that recently.

In early May, eBay announced that it had acquired Expertmaker, a Swedish-based company founded in 2006 that specializes in “providing intelligent solutions powered by artificial intelligence, machine learning and Big Data analytics.”

Expertmaker and eBay had been partners since 2010 on a structured data initiative, but eBay acquired the company to fold it into its existing technology development team.

For a company with a lot of “messy” data like eBay has surely collected on its customers over the years, you can probably see the appeal of developing an AI-based algorithm that can effectively and efficiently sort, compartmentalize and monetize that information.

EBay also acquired SalesPredict, a startup with a learning engine geared for commerce that scans the history of what a shopper has bought to make suggestions about the next thing a shopper might want to buy, in July.

Over the next three or four years, eBay hopes to develop algorithms that better understand the intent of its users and create an “ecosystem involving social networks and messaging applications to allow for better targeting,” R.W. Baird Analyst Colin Sebastian wrote in an investment note in June.

Also in June, Wenig attended the Code Conference in Rancho Palos Verdes, California, and told attendees of eBay’s plans to position itself as a data business rather than an eCommerce business and that developing smart AIs that can help it better analyze data will be vital to the company’s future.

“In the last year, we’ve gotten really serious,” Wenig told Fortune. “I would argue it’s more important for us than our competitors because of the breadth of our inventory. That’s why we’re marching down building the world’s biggest catalog — because we need to have visibility into that inventory. Search is a tool that has marginal diminishing returns when you have that large of an inventory.”

EBay has already begun to “customize” its online marketplace so it’s more of a “curation” of the tastes and shopping habits of each of its users, as opposed to seeing the “same thing as 160 million other people,” according to Wenig.

For the future, eBay seems to be placing a serious bet that artificial intelligence will help turn around its struggling business model by being able to predict what consumers might want to buy on the site and offering it to them more easily.