EBay releases Image Search API to developers

Dive Brief:

Online giant eBay announced it is releasing an Image Search API that will allow external developers to use eBay's image recognition technology for their own uses to help drive more listing, selling and buying activity through the online marketplace, according to an eBay blog post.
EBay also said that Wikibuy, developer of a mobile app and Google Chrome extension for making quick price comparisons, is among the first to use the Image Search API to create a new "eBay visual search feature" for its mobile app.
In addition to Image Search API, eBay has unveiled a Machine Translation API that enables English-to-Mandarin machine-driven translations for item title and description, and Mandarin-to-English translations for search queries. The company has also created Feed Platform for Buy APIs, which lets developers curate, mirror and surface eBay inventory in large volumes.

Dive Insight:

These moves, announced at the eBay Connect developers Conference in San Jose, California, this week, come months after eBay said it would focus heavily on technology efforts, including computer vision and search capabilities, throughout 2018. The company has also been trying harder to work more closely with its community of external developers, with the belief that they can help the marketplace broaden its reach and bring in sales it might not have otherwise closed.

As Gail Frederick, general manager of eBay Portland and vice president of the Developer Ecosystem, put it in the blog post, "eBay's success depends on our talented developer community building quality applications for buying and selling."

That was evident even before this week's news. In the first half of 2018, external developers used public sell APIs to create more than 2.1 billion new listings, manage about four times that number of listings, and drive more than $9.6 billion in gross merchandise volume worldwide. Meanwhile, the company said that public buy APIs were leveraged to drive 1.7 million transactions and more than $85 million in gross merchandise bought worldwide.

Now, giving those developers access to the technology tools used to build features otherwise available only on the core eBay site should help further. These APIs give developers the pieces to build features already on eBay, including Image Search and the related Find It On eBay, both of which the marketplace launched in the summer of 2017, along with the drag-and-drop visual shopping.

Beyond the Image Search API, the machine translation service API should boost eBay's activity with Chinese buyers and sellers — likely a move to combat Alibaba, which is incredibly popular with Chinese consumers in the same space. The company also said it is planning similar Machine Translation APIs to support translation for other languages.

In addition, the Feed Platform for Buy APIs means that developers can store and synchronize millions of items from eBay's site on their own sites, so shoppers encountering the items though those external sites never need to question availability.

EBay also announced this week that it is making all of its Buy APIs — Feed, Search, Item, Checkout, Order and Marketing — available in the U.K. and Germany, a year after doing so in the U.S. Elsewhere, an Inventory API will help developers move a seller's entire catalog of listings to eBay's Inventory model at the same time, rather than having to do it in multiple small batches. There are even more new API efforts mentioned in the blog post.

EBay is under pressure as the number of online marketplaces grows, and Amazon casts a long shadow over its native sector. Nevertheless, the company seems to realize that in order to compete, it needs to spread its technology tools as far and wide as possible.