In nearly all industries, businesses have found ways of implementing generative AI in a way that improves the lives of their employees or meets the needs of their customers better. E-commerce is no different, as eBay recently announced a beta plugin, for instance, that enables sellers on their platform to auto-generate item descriptions based on the product’s attributes using generative AI. Eddie Garcia, eBay’s Chief Product Officer, said the sellers’ feedback on the tool has been very positive and the company is tracking a variety of metrics to measure its effectiveness for sellers and for eBay.
eBay Unlocks Customer-Facing Value with Generative AI
Although this is eBay’s first foray into generative AI, the company isn’t new to AI in general. Up to this point, its implementations have been limited to behind-the-scenes tasks, like fraud detection, recommendation engines, data analysis, and predictive capabilities. Generative AI, however, is unlocking value on the customer-facing side of the platform.
During the VentureBeat 2022 Transform event, Stephanie Moyerman, the former senior director of risk and trust science at eBay, said the company “wants to integrate AI as part of the natural buying and selling experiences and flows… We want to give AI tools to our sellers so that they can enable different, customized experiences for their buyers without having any impedance there.” Her broad language leaves a lot of room for imagination.
Generative AI Use Cases for eBay Buyers and Sellers
While the company’s plugin is in its very early stages, it could easily evolve with more powerful innovations.
For example, imagine if sellers had the ability to take pictures of their item with their phone to generate a set of high-quality images from different angles as generated by AI, paired with a full-fledged, generated description. Or eBay could use computer vision to recognize products as sellers upload pictures of them, scrape data about that product from the Internet, and auto-generate responses to questions that potential buyers might have about that specific product, like ‘What are the dimensions and would it fit in my trunk?’ or ‘Does this model have wireless functionality?’
Which companies are the most important vendors in AI and hyperautomation? Check out the Acceleration Economy AI/Hyperautomation Top 10 Shortlist.
As a two-sided marketplace, eBay must consider the needs of both buyers and the sellers. While the value to sellers of auto-generating item descriptions is straightforward, there is also implicit value for the buyers, who receive higher-quality item pictures and descriptions. Down the line, eBay may also allow buyers to navigate their product searches using natural language rather than filters. Imagine having a conversation about your needs and eBay recommending products that solve that need.
These use cases will also become more powerful as eBay trains models on its own data sets as accumulated through years of buying and selling. eBay can assist in generating product descriptions and product images in a way that’s optimized for conversion and buyer satisfaction.
Embedding generative AI can have a meaningful impact on eBay’s business. Garcia mentioned tracking a number of metrics that the plugin will likely impact, including the sellers’ acceptance rate of generated descriptions, how often descriptions are edited, the conversion rate of products whose descriptions were auto-generated, buyer satisfaction with those products, and the return rates of those products.
eBay’s announcement reminded me of the early days of Airbnb, which hired scores of photographers to take professional pictures of rental accommodations to help early users feel more comfortable about the properties.
In marketplace models, it pays — both the seller and the company — to have good listings. This extends to social media as well, which may assist creators by generating better thumbnails, descriptions, or captions. Generative AI is redefining how such marketplaces are set up, and I’m excited to see how eBay and similar companies move forward.