eBay has allowed prospective buyers to contact a seller with questions about a product for a long time, and now buyers can post a question about a product for other buyers to answer. The feature manifests in the form of a Q&A section on a listing, where you can look at questions from fellow prospective buyers of the same product, as well as answer any questions about products that you may be knowledgeable about. On top of allowing buyers to help out their fellow users, the feature has an AI backend that uses machine learning to compile product knowledge and help users to find others who may have relevant knowledge when they have a question about a product.
While eBay getting the ability to have buyers answer each others’ questions is newsworthy in itself, the real game changer here is bound to be the machine learning AI sitting behind the scenes. As more and more people ask and answer questions about more and more products, the AI will start to learn what types of questions are the most useful with certain types of products, and getting experts involved to answer those kinds of questions early on. Additionally, the AI is likely to help decrease the amount of duplicate questions, as well as unify Q&A sections across different listings that are for the same product or a variant thereof, but are different enough that existing systems don’t catch them, or sellers don’t think to classify them completely. Used smartphones and computers are actually very good examples of exactly this sort of thing.
The new feature is live in the US, but only for desktop and mobile browser users. eBay confirmed that the new feature would be coming to its mobile app “soon”, but did not specify when that could be. Releasing the feature outside of the US, on the other hand, was not mentioned at all, and will likely happen further out in the future, if it happens at all. Given the feature’s AI backend, expansion into other regions and languages would mean that the AI backend would need to be proficient in natural language processing, on-the-spot and contextual translation, and of course, its core function of finding and compiling relevant information.
