Abof, Aditya Birla Group’s online e-commerce portal, has integrated an AI-based system that uses visual imagery to personalise the experience for its customers. This enables them to view products they are more likely to purchase.
This simple image mapping technology has helped the fashion portal to achieve a 95% accuracy rate. The fashion portal is presently integrating AI technology provided by Streamoid Technologies. The tech player offers a computer vision based AI engine.
Prashant Gupta, the CEO of Abof, said, “In the 9 months since we have been live, we are working on improving the whole theme of personalisation. We ask our users to take a simple fashion personality quiz when they sign up.”
“Using their gender and their behaviour during the first few visits we personalise their fashion content feed, that’s how we started out. Now we have moved on to integrating visual-imagery tech through an AI-based system,” Gupta added.
The usual practice that most online fashion e-commerce portal follows is to push recommendations under ‘similar styles’ section using product attributes such as colour and a subcategory such as a dress or a type of dress.
The recommendations are not always accurate or relevant and often result in lower conversions owing to the broad range of styles and colours within subcategories, Gupta claims.
The platform offers two kinds of recommendations on the product detail page, similar styles and completes the look.
