As part of eBay’s dogged pursuit to position itself as a digital shopping destination, the e-commerce company is using personalized data to create a user experience that better appeals to shoppers.

The retailer debuted a personalized shopping homepage today called eBay Interests, allowing shoppers to tailor recommendations on the desktop site and mobile app based on their personal preferences. Users can opt in to the feature by answering a series of questions, such as “What do you love?” and “What are you a fan of?” and marking specific interest areas that help eBay curate personalized items across a broad range of categories. These items will be featured on the user’s homepage, where they will be repopulated with similar items as they sell out.

While personalization efforts are nothing new — retailers have been clamoring to use data to market to consumers and improve online experiences for several years now — the effort is part of an ongoing “structured data journey” for eBay, said Kari Ramirez, communications manager at eBay. Though eBay differs from competitors like Amazon and Walmart since it has no inventory and functions as a peer-to-peer seller, it is still well behind marketplaces with similar models like Poshmark and Farfetch in developing personalized feeds. However, given product discovery on eBay can be difficult, with more than 1.1 billion items for sale across a broad swath of categories, eBay Interests is intended to help shoppers navigate the vast selection while also serving up recommendations based on user tastes.

“Unlike other retailers that have inventory on their shelves they’re trying to get rid of, which they’re merchandising to multiple people even though they’re likely to have different passions and interests, we’re very agnostic,” Ramirez said. “We don’t hold any inventory which makes us very different from other retailers. We’re trying to connect you to what you’re most interested and passionate about.”

Screen Shot 2018-05-17 at 2.35.32 PM

The questionnaire  and recommendations subsequently shared on eBay Interests

Further, as eBay continues to promote its style and beauty inventory to compete with Amazon’s burgeoning private-label fashion business, Interests will help put fashion and designer brands top of mind, said Kimberly Oliver, head curator of buyer experience at eBay. In the last year, eBay has also embarked on a public relations push to educate users that while eBay used to be the home of the bidding war for second-hand goods, today 80 percent of its inventory is new.

“Interests helps surface things people didn’t necessarily know you could find on eBay,” she said. “Instead of having to wade through every dress out there, you could say you like minimalist fashion, and you’ll see Jil Sander and Celine. You’re going to get the brands that will presumably resonate with you.”

In order to gain a deeper understanding of how to best deliver recommendations, Oliver said the team did extensive research on how competitor platforms serve up suggested content. One model example was Spotify’s Discover Weekly, which takes user data to compile a playlist of songs designed to appeal personally to the listener.

“We didn’t really look at other retailers because we’re not like other retailers, given the scope of our catalog,” Oliver said. “But we looked at things like Spotify, Pinterest and a bit of what Instagram is doing to see how they make recommendations that are a bit more human and a bit more intuitive. We want to be the Spotify of stuff.”

Though eBay has experimented with curation in the past, tapping fashion editors to create lists of products featured on the site, the turnover rate was so fast that most of the items would sell out by the time a reader saw the post. To combat this, the algorithm functions by setting up a baseline of preferences that rotates products in and out as they are available, anticipating for a user’s lifestyle, tastes and proclivities.

Just like any algorithm, it gets smarter the more a shopper uses it and feeds it data in the form of search terms, creating an even more accurate portrait of the consumer over time. Interests is also attuned to regional tastes and is personalized based on user information in different parts of the world. For example, Oliver said users indicating an interest in sports in the U.K. will see more cricket-related posts than in the U.S., where fewer enthusiasts of the sport live.

“This is curation at scale,” Ramirez said. “Over time, the homepage that you see might look different based on your life and milestones like weddings that yield bridal couture or baby items. Our interests are definitely not static. With life changes, Interests will change with you, as well.”