- Don-Alvin Adegeest |
Visual search, while hardly a novel concept, is quickly becoming a tool for e-commerce platforms to allow users and customers to search for what they want using images.
eBay this week announced it has expanded its image search feature with the possibility to drag and drop images into its search bar while users are using the platform to shop on their mobile app.
The idea is that similar items can be more easily found in its marketplace, something a key word or hashtag may not necessarily locate.
According to a company statement, eBay claims its new new visual shopping feature is powered by AI and uses deep learning networks known as convolutional neural networks to process your images.
When a potential customer submits an image in the search field, the neural network converts it into a vector representation. Then, the vector representation of the image that you submit is compared against more than 1.1 billion live listings in eBay’s marketplace, using nearest neighbor search. eBay then displays the best-matched items, ranked by visual similarity.
While we rarely think of shopping as visual searching, that is the reality of how consumers shop: we see an image and click on it. Imagine your favorite online store without images of the garments and accessories it is selling? Would you purchase a dress or pair of jeans based solely on a description of the item? Likely not.
Visual content is making its way to traditionally text-based search, especially when it comes to online shopping. According to a recent study by eMarketer, 72 percent of internet users in the US regularly or always search for visual content prior to making a purchase.
Credit: eBay visual search