GenAI Helping Amazon Sellers Cut Costs, Improve Ad Bidding: Machine Learning VP

A growing number of Amazon merchant sellers are using generative artificial intelligence tools to list their products on Amazon, as well as to bid for advertisements on the platform, according to the e-commerce giant’s Vice President, Machine Learning, Rajeev Rastogi. He told Outlook Business that the increasing number of users gives the company confidence that these tools are delivering real value.

The online marketplace released new GenAI features in 2024 that allow sellers to pull all product information by simply copying and pasting a link.

Last December, the platform announced plans to support more than 15 million small businesses through AI-enabled tools integrated across its ecosystem. These include agentic AI seller assistants, upgraded Seller Central capabilities, generative AI listing tools, a creative studio for ad development and a video generator aimed at reducing the cost of video advertising.

"Sellers appreciate them (GenAI tools), as they may otherwise have to hire consultants, which would add to their costs. These tools not only help them save on such fees but also enable more accurate bidding and budgeting decisions," said Rastogi.

However, he did not disclose the exact number of sellers using these tools.

Amazon is leveraging GenAI to address one of the key pain points sellers face, according to its Machine Learning VP.

"Earlier, the process (of listing products on Amazon) was quite cumbersome, as sellers had to manually fill in multiple attributes and product details. With GenAI, we’ve significantly streamlined the entire product listing process," he said.

Explaining further, he said that sellers now only need to upload an image along with a few descriptive keywords, and the system can automatically extract attribute values and populate a large percentage of the required fields.

"This substantially reduces the workload and effort involved for sellers," he said.

On the advertising side as well, GenAI tools recommend which keywords sellers should bid on and what the appropriate bid values should be, based on their objectives.

"For instance, if a seller wants to adopt a more aggressive sales strategy, they can choose a corresponding bidding approach. Alternatively, if they prefer to be more conservative and reduce advertising costs while aiming for moderate sales, the system can recommend suitable bid values accordingly," Rastogi said.