After recently adopting generative AI to elevate its product reviews, e-commerce leader Amazon has unveiled new AI-driven features to streamline the online apparel shopping experience. The company now employs large language models, generative AI, and machine learning to enhance four key functionalities aimed at helping customers find the right clothing fit—a persistent challenge in online shopping and a primary reason for high apparel return rates.
According to Coresight Research, the average return rate for clothing purchased online stands at 24.4%, significantly higher than the overall online return rate by eight percentage points. Retailers have noted that online returns have surged over the past two years due to consumers purchasing multiple sizes or colors, often returning what doesn’t fit as the ease of home try-ons and return shipping has improved.
To combat this issue, Amazon has integrated AI into the online shopping experience through four innovative features: personalized size recommendations, a "Fit Insights" tool for sellers, AI-curated highlights from customer fit reviews, and redesigned size charts.
With personalized size recommendations, Amazon Fashion utilizes AI to create a deep learning algorithm that recommends the best-fitting size across different styles. This system analyzes the relationship between various brands’ sizing systems, customer reviews, and individual fit preferences to provide real-time size suggestions, adapting to changes in the customer's size needs. However, this feature may face challenges when family members shop for different sizes, such as when parents buy for their teenage children; Amazon advises that distinct profiles can be created through the “find your size” tool to mitigate confusion.
Another innovative feature, “Fit Review Highlights,” aims to provide clarity on fit, addressing potential size confusion. This tool builds on the AI-generated Customer Review Highlights launched in August 2023, which summarizes customer sentiment and key product attributes with clickable buttons for easy navigation.
With Fit Review Highlights, Amazon extracts fit-related details from customer reviews—including size accuracy, fit on specific body types, and fabric stretch. Large language models help summarize these insights into user-friendly highlights tailored to each shopper, saving them time by eliminating the need to sift through hundreds of reviews.
Amazon is also enhancing size charts throughout the site. Using large language models, Amazon Fashion streamlines product size data from multiple sources into standardized measurements, correcting inaccuracies and ensuring more precise charts for better fit.
Sellers benefit as well from AI-powered insights via the Fit Insights Tool, which provides an overview of customer fit needs, enabling better communication about sizing—such as clarifying whether items run true to size or larger/smaller. This aggregated data not only enhances customer interactions but also informs sellers’ future production strategies, as large language models analyze customer feedback, returns, and size chart discrepancies.
These advancements are just a glimpse of how Amazon is harnessing AI to enhance the shopping experience. In addition to customer review highlights, Amazon has introduced generative AI tools for sellers to improve product descriptions and optimize product images, which could increase click-through rates by an estimated 40%. Beyond its e-commerce platform, Amazon is implementing AI across various consumer products, including Alexa and Fire TV.
Ranking
