Amazon launches AI tool ‘Help Me Decide’ to guide shoppers

Amazon has launched a new AI-powered feature called Help Me Decide, designed to recommend products to shoppers based on their browsing and purchase history.

The feature is available to millions of US customers using the Amazon Shopping app or mobile browser.

Help Me Decide analyzes viewed items, search activity, and past purchases to suggest a product, and also provides upgrade and budget alternatives.

The tool uses large language models and Amazon’s own cloud services to match recommendations with customer needs and preferences.

Users can access Help Me Decide by tapping the button on product pages after browsing several similar items, or through the “Keep shopping for” section on the app’s homepage.

Amazon said the feature aims to make it easier for customers to choose between similar products.

🔗 Source: Amazon

🧠 Food for thought

Implications, context, and why it matters.

Amazon’s AI shopping tools lack public performance metrics, limiting investor visibility

Amazon announced Help Me Decide as part of its AI shopping suite. The Q2 2025 earnings call mentioned no conversion rates, average order values, or impact from these features 1.
AWS’s AI business grew at triple-digit rates, and management gave numbers on supply limits 1. Retail tools like Rufus, Amazon Lens Live, and Help Me Decide had no metrics in Q2 2025 1.
The stores business (Amazon’s retail segment) focused on metrics like premium brand selection, everyday essentials growth, and faster delivery 1. Everyday essentials grew faster than the rest worldwide and made up about one of every three units sold 1. AI shopping help was not named as a driver, which implies early adoption or no quantified impact yet.

Third-party sellers can lift placement in LLM product picks

Help Me Decide uses product details and customer reviews to rank options. Sellers can tune attributes, A+ content (enhanced product detail pages with rich media plus comparison modules), and reviews to surface in AI suggestions.
Amazon raised required product attributes to 274 across 200 types as of August 2023 2. These structured fields include size and color. They also include material and wattage, plus compatibility. The company updated the rules several times since then, including September 2023 3 and later changes 4.
Software as a Service (SaaS) tools that keep listings compliant and readable for LLMs could see demand as AI shopping reaches more customers.