Generative AI Will Now Help Write The Listings You See On Amazon

If Amazon is focused on anything, it would be zooming products to our doors as quickly and effortlessly as possible on the buyer's part, and in ludicrous quantities. Statista notes that $356.11 billion worth of products were bought on Amazon in the U.S. in 2022. However, Amazon isn't solely about convenience (or intrusion, depending on your perspective) for buyers. Those buyers, in lots of cases, are reliant on third-party sellers.

On September 13, 2023, the company announced that life was about to become far easier for those sellers, too. The best way to ensure potential customers want to buy your product is by providing in-depth description, and this can be a painstaking process. 

This is why Amazon is introducing generative AI tools designed specifically for filling out product descriptions, tasked with "recognize, summarize, translate, predict, and generate text and other content, to build more comprehensive product descriptions." Here's how the controversial might of generative AI is going to be employed in Amazon product listings.

Vice President of the Worldwide Selling Partner Experience at Amazon, Mary Beth Westmoreland, has introduced how the new tool will work in the company's small business news blog. In short, a large language model will do what AI does best: Use the comprehensive array of samples of data to create lengthy and valuable product listings from basic information provided by the seller.

In the example provided by Amazon, the user navigates to the Product Identity tab and simply inputs "Amazon Basics mousepad with gel wrist pad" into the "Generate Listing Content" box. From this, the product description and Bullet Point summary boxes are filled in with much more detail about the product. For example, the description states that the mousepad's "smooth, consistent surface is perfect for both laser and optical mice" and that it is "easy to clean with a damp cloth."

It's prominently noted within Seller Central that this feature is in early access, and that users must vigilantly read through and edit the content to ensure its accuracy — and eligibility with Amazon's guidelines for sellers in mind. Fortunately, users can freely edit any of the resulting text.

Amazon Select and Catalog Systems vice president, Robert Tekiela, is quoted in the blog post boasting that the system "can infer a table is round if specifications list a diameter or infer the collar style of a shirt from its image," when creating descriptions. Tekiela also made sure to highlight another fundamental truth of generative AI: it has enormous potential and utility when used carefully and conscientiously.

The bottom line is that the global powerhouse has a tremendous responsibility to its third-party sellers (which Amazon notes account for around 60% of its sales), and any moves that may make their work more convenient are very welcome indeed.