AI Size advisor
A Size advisor that suggests the best fit for users when they are buying clothes in a marketplace
00
problem
42% of online shoppers say they aren't sure what size to buy. 40% of online purchases are returned, and over 58% are due to the wrong size or fit. The fashion industry is the second most polluting industry worldwide.
solution
Companies that can suggest to customers the right size increase sales by 15%. Help customers find the right size and reduce returns (and increase sales while you are at it). Help reduce CO2 impact by reducing returns (which lead to CO2 emissions and clothes going to landfill).
The project combined artificial intelligence and computer vision to create a simple photo-based sizing tool. Users take two photos following guided instructions, and the AI analyzes their measurements to recommend the best size for each garment, addressing the fundamental challenge of online clothing fit.
I used a user-centered design approach combining technical feasibility research with extensive user testing to create an intuitive interface for complex AI technology. The approach started with desk research on similar platforms and user pain points around online sizing. I conducted affinity mapping to organize findings and built information architecture for various user scenarios. The focus was on making sophisticated AI technology accessible through clear visual guidance and step-by-step instructions.
year
2019
timeframe
4 weeks
tools
Figma, Hotjar & Google Analytics
category
AI strategy
01
Illustrations showing the step-by-step instructions helps users to understand what to do
02
Simplifying the flow by asking the most essential information for users
03
The result suggest users which could be the perfe fit according with the two pictures they take
outcomes
The first Marketplace who tested eStylar feature was Draper James who faced high return rates due to poor fit, leading to customer frustration and increased costs. The solution uses AI and computer vision to help shoppers find items that fit their exact size and body shape. After one week users
learnings
Working with a few amount of outsources force you to be more creative. User testing and tracking after implementation were a key for the impovement of the project.
55%
Adoption rate
164 of 300 users engaged with the size advisor
8%
Convertion rate
8 users purchased the product based on the AI recomendation
5%
Product satisfaction
Only 1 persona return the item of 8 purchases
