Target knows you are pregnant. Homegrown e-commerce firm Flipkart wants to know if you are a doting father. That is, not in a creepy way, they say.
At Flipkart, which hit $1 bn in gross merchandise value a full year before the target, all the talk is about building intelligence into its systems, i.e. going beyond scale.
Last week at the company’s annual hack day held at the Indian Institute of Science, its CEO Sachin Bansal said that now its focus will shift “from scale to intelligence.”
— Flipkart (@Flipkart) March 7, 2014
Armed with data generated by over 14 million odd registered users on Flipkart, the idea is to know each user to such an extent that they end up being shown the most relevant products on the website.
“If we know that its a father who is on the site, and a doting father at that, we’d love to show him what he is looking for,” says Amod Malviya, the head of engineering at Flipkart (Read: Engineering the Flipkart story).
The e-commerce firm’s growing mobile user base is also making it easier for the company to personalize the shopping experience. Each mobile phone has a unique fingerprint that can be mapped to a user.
“Increasingly, across industries, systems are becoming intelligent. Where systems learn the intent of the customer,” says Malviya. For instance, a user could be looking for Harry Potter, the movie, the game, or merchandise. “When my wife is searching, its probably the book. My son, is looking for the merchandise,” he says.
The company has also started setting up a data science team to play with such problems and come out with solutions that have a direct impact on sales.
Some of the questions the team will ask are: How do you personalise for every single aspect of the shopping experience? What are the techniques useful in doing intelligent pricing? How does one plan inventory specific to the Indian context?
Flipkart is also experimenting with search techniques that go beyond simple semantic search. For instance, if someone is looking for a “red dress,” the shopping site wants to be able to show all red dresses, even if it hasn’t been textually tagged a “red dress.” This feature is currently being tested on the site.
Amazon and other global e-commerce players have already cracked personalization in a big way. E-commerce companies and brick and mortar stores go great lengths to understand their customers better by using different means. Some such practices, however, are not always well received. For instance, the practice of using browser cookies to track a user’s behavior, although one of the oldest trick in the book, doesn’t go very well with users who wish to have greater privacy on the Internet.
Right from data like the customer’s location, or the time of the day when he is using an app Flipkart is interested in every data point, “so long as customers privacy is not breached.”
Malviya says that the company is “maniacal” about customer’s privacy and data security.
Image Credit: Shutterstock