The news: Amazon has begun rolling out its AI-powered shopping assistant Rufus to mobile app users.

• Because the tool is trained on the retailer’s product catalog, customer reviews, community Q&As, and information from across the internet, it can respond to a broad array of questions tied to consumers’ specific needs or purposes, compare one product to another, or assist in product research.
• For example, a shopper might ask queries such as, “What gear do I need to go camping in the fall?” or “what to consider when buying headphones.”

How it works: Shoppers can ask Rufus questions by typing or speaking into the search bar in Amazon’s mobile app.

• They can expand the chat layer to see answers to their queries, tap on suggested questions, and/or ask follow-up questions.
• They can return to traditional search results by swiping down or tapping outside the chat layer.

Why it matters: While Amazon’s vast product catalog and effective search tool make it easy to quickly find and buy an item such as Tylenol PM or AAA batteries, it also makes it challenging to use the platform to browse or seek inspiration.

• Rufus attempts to solve some of those problems by culling through Amazon to surface relevant information and results.
• However, the reality is that in our early testing, results were often hit or miss; sometimes we received a note that said no results were found and we should check our spelling or use more general terms, and other times results were irrelevant to our search.
• An Amazon representative tells us that it is still early days for Rufus, which continues to improve its model and fine-tune its responses.

The big takeaway: Rufus has great promise that could vastly expand the amount of time consumers spend on Amazon, which could boost sales and retail media revenues. But that will only occur if it can quickly improve the quality of its responses.