Skip to Content

Walmart’s New High-Tech Labs: You’re Not in Arkansas Anymore

The way we shop is changing at lightning speed, and the world’s largest retailer knows it needs to keep pace.
October 16, 2012

In the lobby of @WalmartLabs, as this outwardly drab office building in San Bruno, California, is known, there’s a sign informing visitors they’re exactly 1,849 miles from Walmart’s headquarters in Bentonville, Arkansas. The sign hints at the vibe within. This Silicon Valley innovation lab feels more like an outpost of Google or Facebook than like any part of a global retail corporation that employs 2.2 million associates at nearly 10,000 stores around the globe.

Big business: A sign in the lobby at Walmart Labs.

Inside the new offices, built in the last 18 months, Internet industry veterans have organized the staff into small, tight-knit, fast-moving teams that run independent projects, interrupted once in a while by activities like ping-pong tournaments, catered Friday lunches, and a hack day earlier this year. But the casual feel belies a serious mission: to develop, from scratch, “core technology” that will bring Walmart into the Internet age.

Walmart Labs is a crucial part of Walmart’s plan to expand its Internet retail business and catch up to e-commerce giants like Amazon. With more people buying on the Web and using smartphone apps to make lists, find deals, and comparison-shop, the world’s largest retailer knows it needs a broader online reach. It also is under pressure to figure out the best ways to use mobile and social-networking technologies in physical stores as the differences between e-commerce companies and physical retailers start to blur.

On the Web, where Amazon dominates with nearly $50 billion in total sales compared with Walmart’s target of $9 billion for global e-commerce next year, Walmart wants not just to catch up but to leapfrog ahead. In August, it became the first big-box retailer to launch a homegrown search engine, called Polaris, rather than use technology from a vendor. Polaris uses semantic search algorithms that can understand some of the meaning behind a search query. For example, it can determine that a woman typing “flat” might want shoes, not a flat-screen television. The company claims the new engine has boosted search-related online sales by 10 to 15 percent

In other areas, staff are working to develop personalized homepage recommendations and to help Walmart store managers use Twitter trends to inform decisions about products to stock. The company also seeks to combine online data with information gleaned from real-world shoppers in its many stores across the globe.

The nucleus of Walmart Labs, formed in April of last year, consists of around 50 people acquired along with the startup Kosmix, which developed natural-language algorithms that improve and personalize search results by categorizing words and understanding them in context. Sri Subramaniam, a Walmart Labs vice president who started with Kosmix, compares their project to Google’s new “Knowledge Graph.”

Though both Kosmix’s founders left this summer, Walmart has put their technology to use in the new search engine and expanded the labs to include 400 staff members located in California and in Bangalore, India.

Chris Bolte, a Walmart Labs vice president who heads a team working on user acquisition, says Walmart’s advantage over many online companies lies in the sheer scale of its real-world operation. When combined with online information about customers, data on physical retail purchases can provide “a huge look at what is happening in our economy,” he says. Last week, Bolte was about to run a “small” test on a new tweak of the e-mail algorithm involving millions of customer accounts. It would begin taking a recipient’s local weather forecast into account when making purchasing suggestions. Buying habits often change on gloomy days, the staff behavioral scientist believes.

It’s clear, however, that there’s a limit to how much commerce will move online. For instance, there’s no substitute for trying on a pair of pants in person, or leaving a store with tonight’s dinner in your hands.

And so Walmart is looking to shape the future of commerce not only on websites and social networks but also in places like a newly opened store in nearby San Jose, where Walmart Labs staff has been given relative freedom to experiment with the realm where online and offline shopping meet. More and more, people research purchases on the Web, take suggestions from Facebook friends, and use smartphone apps to make shopping lists before going to a real store.

Mobile technologies are already sharpening the competition between online and physical retailers. This year, Walmart kicked Amazon’s Kindle tablets off its store shelves over concerns that users would use the device to search for cheaper deals online.

Last week, Walmart announced it would join Amazon in experimenting with same-day delivery for online purchases. In Walmart’s case, its stores can become distribution nodes. One technology being tested at the San Jose store is an app called “Endless Aisle,” which allows shoppers to scan a code for out-of-stock items and see similar inventory to order online.

Subramaniam credits Walmart for maintaining the startup culture after the acquisition. A few weeks ago, with the coöperation of Walmart lawyers, he took a technology research project started in the Kosmix days—it processes fast-moving data from Twitter—and released it to the open-source community.

Approving more open-source projects could help Walmart maintain credibility as it seeks to compete with other Silicon Valley technology companies for talented engineers. Even in this effort, Walmart is trying to emphasize the scale of its existing business. A billboard on Highway 101, a main drag in Silicon Valley, brags: “Changing the way people shop. 1,000,000,000 at a time.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.