Problem / Objective:

Predicting upcoming fashion trends is one of the biggest challenges the e-commerce industry is facing today. Despite having a large amount of data, these companies fail in understanding customers’ preferences and what they are more likely to purchase. 

Generally, companies spend months speculating on what customers would like to wear, and then prepare it and introduce in the market -which can take a few months. 

Solution / Approach:

To overcome the time lag caused in predicting fashion trends, e-commerce platform Myntra, started using AI. They started by scanning the entire digital space, including fashion portals, social media, and Myntra customer data. By using computer vision and machine learning on the scanned data, the platform was able to figure out what customers love. These insights are further passed on to the company’s lab for implementation in designs, which puts the finished products on the Myntra app.

Data on whether customers are buying products or not are fed into a computer, which continues to learn and throw up what works best for customers. The technology behind this platform uses collaborative filtering on the demographic data which is collected from the company’s database. The metadata associated with the demographic data allows recommendations of clothing such as the design, printing patterns, size, etc.

Impact / Implementation:

By using AI in e-commerce, Myntra was able to bring out collections much faster. Currently, other brands sell 50 % of their new collections over a period of 4-6 months at full price, and the rest is sold at a discounted price. The use of such technologies can help brands to sell 80% of their new inventories at full price in just two months.