Data Science Congress

Data Science in eCommerce and Retail

The retail industry has steadily warmed up to eCommerce in the last two decades. It all started with the dotcom boom in 1999, when usage and adoption of internet by businesses and consumers was over-speculated. This followed by a massive bust in 2001. The industry realized that survival of eCommerce companies was dependent on, more than anything else, robust engineering capabilities. Technology became the key differentiator for companies to survive and excel in the e-commerce industry.

As Dr. Ravi Vijayaraghavan, Vice President and Head of Analytics at Flipkart stated “the only way to understand customers in eCommerce industry is through data“.

This constraint mandated companies to build strong data capabilities to understand customer behavior, buying patterns, sales performance etc. The strong data capabilities also turned out to become the greatest advantage for eCommerce companies to penetrate and scale in the retail markets. With other factors contributing to the ecosystem, like increased internet usage, evolution of smart phones and advancements in technologies, new global e-commerce giants like amazon and eBay emerged. These companies also forced offline retail giants to enter into the eCommerce space.

The data science capabilities within eCommerce industries helped companies to get insights that would have never been possible otherwise. Dr. Ravi Vijayaraghavan mentioned some of the interesting industry insights that Flipkart found out using their data science capabilities, when he was speaking at the data science summit 2017, organized by the data science congress in Mumbai, India.

18 Ravi Vijayaraghavanpanel discussion - Retail

When analyzing factors that drive sales from women customers, he said that the data scientists at Flipkart found a correlation between number of swipes made on a product with higher sales numbers, highlighting the fact that products offering multiple choices sell better with women customers. As a contrast, they found that branded products sold better with men.

He mentioned that another important application of data science is identifying frauds where sellers pose as users, order products online and return them giving bad ratings. Flipkart uses data science to identify such sellers by analyzing their behavior patterns.

Flipkart also uses data science techniques to improve product images by removing unnecessary shadows or highlighting certain parts of the product image. Dr Vijayraghavan said he would further like to use data science capabilities to understand new markets and improve Flipkart’s penetration into tier 2 & tier 3 cities, where people dont necessarily have access to good quality goods and products.

Bhanu Ramgopal, VP of Engineering and Analytics at Myntra, an online shopping site in India for Fashion, said that they were using data science to improve product recommendations to customers, in order to up-sale & cross sale their products. They are also thinking of providing to their customers, the facility to order a variety of products so that they can actually try them out at home, thereby giving them an ‘in-the-shop’ experience.

Piyush Chowhan, VP & CIO of Arvind Lifestyle brand, talked about integrating offline retail store data with online data to build a truly customer centric organization. Arvind Lifestyle brand is India’s largest denim manufacturer and world’s fourth-largest producer and exporter of denim. It has a huge offline nationwide retail network of fashion apparels and is a licensed retailer of many international fashion apparel brands.

20 Piyush Chowhan

Mr Piyush Chowhan said that retail fashion industry has been undergoing tremendous disruption in the recent days. He said that data science can provide insights about fashion trends more accurately than vogue fashion weeks in London and Paris, as it gives a true reflection of buying patterns. He also pointed out that stable fashion choices are emerging. According to Mr Piyush Chowhan, understanding why a customer comes to the store and leaves without buying anything is an extremely important use case of data science in both online as well as offline retail industry.

The data science summit 2017, organized by Data science congress, was attended by various high profile corporate and political delegates across the country and many data science professionals, students and enthusiasts. Data science congress is trying to build an ecosystem which will help India to become the center stage for the research and development of data science skills. The event was organised by Aegis School of Business, Data Science & Telecommunication.


Ninad Joshi has 18 years of work experience in technology & management consulting. He has wide exposure in the field of data science, data engineering and data analytics, and has in-depth knowledge of industry standard solutions and service offerings within the sector. Email: Mobile: +91 7045 290 284 Location: Mumbai, India