Walmart announced the launch of WalmartLabs in India today, which focuses on the intersection of social, mobile and retail. Anand Rajaraman, SVP at Walmart Global Commerce, speaking at the NASSCOM Product Conclave in Bangalore, said that in the last two-three years, there has been a fundamental change in consumer behavior, in terms of the importance of mobile and social, and for the first time, companies have access to information outside of consumer behavior on their own sites. Rajaraman was previously with Amazon, joining the company after, a company he co-founded, was acquired by Amazon. Notes from his talk:

Social & Mobile Is Important: these are fundamental and massive shifts. about two quarters back, the number of smartphones went ahead of feature phones. People are spending more and more time on social media.

Making Sense Of Unstructured Data On Social Networks: One of the key realizations that we had on amazon was that we can use data to improve customer experience. If you’ve used Amazon, you’ll have people who bought this product also bought the other one. The interesting thing is that the data that most companies had access to was gathered at the company itself. I can use this information to make the site better. The change now is that there is massive data on Facebook and Twitter, much more than on my site. So Far, we’ve had data on how customers interact with us, but not what they’re doing when they’re not interactive with us.

All this data, to the first approximation, is available, through a Facebook app. What we’re doing is that we’re trying to use this data which is publicly available. Big data is a tide that is goig to raise a lot of boats. We’re trying to understand what people’s interests are, and the interest graph interest graph as opposed to the social graph.

What about the noise on social media: The power of tech is to aggregate the data, remove the noise and figure out what people are looking for. You’re building the interest graph on the basis of the social graph.

In-Store Networks: The change is that people are sharing location information. At Walmart, we have 4000 stores all across the world. Each store is in a different community. So far, it has been really hard to understand what the interests of the people in the neighborhood are. We are looking at the peopele who look in the community of each store, and then doing an analysis, and figure out what products are to be carried.

Social & Gifting Information: the other piece is to use info to recommend things to you. We all have friends and we gifts to them. Gift giving is an activity in the holiday season. But how do you know what your friends like. You can see what kind of products and music they like on Facebook.

How can data changes things? We’re familiar with f=ma, which si fundamental to science. These don’t come out of vaccum, they come from data. An astronomer made a detailed record of in the 15th century, and of the position of the stars. Kepler studied that data and came out of laws of planetary motion, and then newton used it for. You start with data.

The biggest datasets have so far come from the skies. For the first time today, we have data about ourselves at unprecedented scale. Social networks give us data on our society. This data is going to revolutionize every field of human endeavor. This is a new gold rush. It’s going to revolutionalize business, but it also going to revolutionize science.

Blurring Lines Between Online & Offline Commerce:  Two fundamental moods of shopping – ecommerce and offlien commerce. What will happen over the next few years is that the lines will blur and you wont know if you’re shopping online or offline. We rely on so much information that we take it for granted. When you walk into a physical store, there are no ratings and reviews in a store. There should be as much information in stores as online. That might come from social media, from what your friends have bought. The first generation of ecommerce was about listing store information on the web. The next generation is to take the information on the web to the store – from the store to the web, to the web to the store.

Info In Store: using info about repeat purposes and providing incentives is going to work. On hyperlocal, one of the more interesting kinds of things that we’re trying out is that when you’re in a store, often a store associate doesn’t have a clue. Sometimes there might be another customer who can help.  Can we build a dynamic social network in the store, and can you ask someone in the store to help with it? it’s hyperlocal to that store.

Patterns In Data: as more and more data becomes available, we need to train people to mine the data and come up with patterns and insight. for the last 4-5 years, i’ve been teaching a class on this topic called data mining. when I first started, in 2004-5, we had 20 students, now we have 120-150. More and more people want to learn how to do this.

People wont find products, products will find people? We find products in many ways, we go looking for them because we’re in discovery mode. If you know exactly what you want, search can find it. The big hole today is the space of discovering things, and there are new products and you don’t even search for them because you don’t know they exist. What form will this take is hard to know.