Finding the perfect fit is hard enough in the physical world, but navigating the virtual space adds another layer of complexity to picking out and purchasing clothes. Amazon is trying to help shoppers overcome these challenges and is channeling efforts into enhancing customer confidence in fashion shopping through AI-fueled innovations.
In a recent blog post, the company outlined four ways it’s leveraging AI – personalized size recommendations, review highlights, re-imagined size charts, and fit insights – to improve the shopper experience.
- Size Recommendations
AI-driven size recommendations are personalized for each customer, removing the need for manual cross-referencing of reviews, size charts, and product details. Amazon’s deep learning-based algorithm takes into account sizing relationships between brands, customer fit preferences, and product details to recommend the best-fitting size in real-time.
The algorithm then clusters customers with similar size and fit preferences and learns from millions of product details and anonymized customer purchases, continuously learning and adapting to changes in customers' size needs accommodating potential size changes over time.
- Fit Review Highlights
Amazon leverages large language models to offer personalized review highlights for each customer, analyzing common themes in reviews related to their recommended size. This approach aims to help customers access personalized size guidance and make informed decisions about how a particular style will fit them. The feature then advises a customer whether to size up or down in a particular style based on reviews from customers who have bought the item in the same size.
- Size Chart Accuracy
Amazon is working to improve its size chart accuracy, tapping LLMs to automatically extract and clean up product size chart data from multiple sources. Amazon then transforms the data into standardized sizes, sifts out duplicate information, and auto-corrects missing or incorrect measurements.
- Fit Insights Tool
The Fit Insights Tool uses a large language model to gather and analyze customer feedback related to fit, style, and fabric in the context of returns and size chart evaluations. It employs machine learning to identify defects in size charts, allowing brands to gain insights into customer fit issues. By understanding this feedback, brands can improve communication of sizing information to customers and roll valuable insights into future design and manufacturing processes.