×

The Amazon-Whole Foods deal is less about stores and more about data, says data scientist

  • Amazon will likely use data from all of its platforms to create an ambient e-commerce experience for customers, said Kenneth Sanford of Dataiku.
  • Instacart is working with traditional retailers to bring them customer data insights.
  • The startup is also experimenting with computer vision to make it easier for its shoppers to find things on shelves.
Amazon President, Chairman and CEO Jeff Bezos.
Mike Segar | Reuters
Amazon President, Chairman and CEO Jeff Bezos.

Amazon's acquisition of specialty grocer Whole Foods is not only about acquiring hundreds of stores and affluent customers — the real value of the deal is in all of that customer data, according to one data expert.

Kenneth Sanford, data scientist at Dataiku, told CNBC recently that one of Amazon's goals should be to combine the data it already collects from its online platforms, Echo and Alexa, with Whole Foods' customer transaction data.

In that regard, the company can predict what customers need and automatically send it to them, creating a personalized "auto grocery" experience, he said. "Amazon will know what's in your refrigerator already and will be able to deliver extra turmeric when you need it," said Sanford.

Essentially, Amazon will leverage its know-how in technology to customize the individual grocery shopping experience — all without the consumer ever even leaving the house.

"They'll be able to say, 'I know you have these three things in your cabinet, I'll send you these two additional things, based on the fact that you haven't had it in three weeks and you liked it last time,'" Sanford said. "They're going to be able to create meal bundles at a level that, say, [the] Blue Apron's of the world could never really do," he added.

Amazon did not immediately respond to CNBC's request for comment.

Grocery buying as a way to build relationships

Ambient e-commerce is something Instacart, one of Amazon's competitors in the grocery space, is already working toward.

"We're really deep into the consumer data, because we have a direct relationship with the consumer," Jeremy Stanley, vice president of data science for Instacart, told CNBC. "One of the wonderful things about groceries is that compared to other e-commerce purchases, groceries are habitual and frequent. People need groceries every week," Stanley said.

Instacart already uses customer behavioral data and search activity to anticipate what a customer wants or might like, he said.

The company is also using machine learning to improve logistics. "We're rapidly experimenting with computer vision for different types of applications, especially in understanding our catalog and enriching our catalog also potentially in identifying items on shelves in stores. There are lots of other directions that we can go with this," Stanley said.

Like Amazon, Instacart's online retail business is built on optimizing customer purchasing, relationships and recommendations by using the most innovative data science, machine learning and artificial intelligence technology.

"It's the digital and end to end asset … being able to understand everything from when the customer first shows up what their intent was to the long tail of their transactions in the months and years to come," Stanley said. "That kind of integrated data doesn't exist in many classic businesses."

The delivery startup is working with traditional grocers to collect data and insights that will allow them to remain competitive.

"Our customers are the customers of the retailers we serve, there is such brand loyalty to the retailers," Stanley said. "Groceries are really personal, maybe even more personal than the movies you watch, and I think data can really change the way people buy food."