Skip to Content

Why Google Doesn’t Have a Research Lab

Google manages to invest in fundamental research without having a traditional lab in which to do it.
March 18, 2014

Research vice presidents at some computing giants, such as Microsoft and IBM, rule over divisions housed in dedicated facilities carefully insulated from the rat race of the main businesses. In contrast, Google’s research boss, Alfred Spector, has a small core team and no department or building to call his own. He spends most of his time roaming the open plan, novelty strewn offices of Google’s product divisions, where the vast majority of its fundamental research takes place.

Groups working on Android or data centers are tasked with pushing the boundaries of computer science while simultaneously running Google’s day-to-day business operations.

“There doesn’t need to be a protective shell around our researchers where they think great thoughts,” says Spector. “It’s a collaborative activity across the organization; talent is distributed everywhere.” He says this approach allows Google make fundamental advances quickly—since its researchers are close to piles of data and opportunities to experiment—and then rapidly turn those advances into products.

In 2012, for example, Google’s mobile products saw a 25 percent drop in speech recognition errors after the company pioneered the use of very large neural networks—aka deep learning (see “Google Puts Its Virtual Brain Technology to Work”).

Alan MacCormack, an adjunct professor at Harvard Business School who studies innovation and product development in the technology sector, says Google’s approach to research helps it deal with a conundrum facing many large companies. “Many firms are trying to balance a corporate strategy that defines who they are in five years with trying to discover new stuff that is unpredictable—this model has allowed them to do both.” Embedding people working on fundamental research into the core business also makes it possible for Google to encourage creative contributions from workers who would typically be far removed from any kind of research and development, adds MacCormack.

Spector even claims that his company’s secretive Google X division, home of Google Glass and the company’s self-driving car project (see “Glass, Darkly” and “Google’s Robot Cars Are Safer Drivers Than You or I”), is a product development shop rather than a research lab, saying that every project there is focused on a marketable end result. “They have pursued an approach like the rest of Google, a mixture of engineering and research [and] putting these things together into prototypes and products,” he says.

Cynthia Wagner Weick, a management professor at University of the Pacific, thinks that Google’s approach stems from its cofounders’ determination to avoid the usual corporate approach of keeping fundamental research isolated. “They are interested in solving major problems, and not just in the IT and communications space,” she says. Weick recently published a paper singling out Google, Edwards Lifescience, and Elon Musk’s companies, Tesla Motors and Space X, as examples of how tech companies can meet short-term needs while also thinking about far-off ideas.

Google can also draw on academia to boost its fundamental research. It spends millions each year on more than 100 research grants to universities and a few dozen PhD fellowships. At any given time it also hosts around 30 academics who “embed” at the company for up to 18 months. But it has lured many leading computing thinkers away from academia in recent years, particularly in artificial intelligence (see “Is Google Cornering the Market on Deep Learning?”). Those that make the switch get to keep publishing academic research while also gaining access to resources, tools and data unavailable inside universities.

Spector argues that it’s increasingly difficult for academic thinkers to independently advance a field like computer science without the involvement of corporations. Access to piles of data and working systems like those of Google is now a requirement to develop and test ideas that can move the discipline forward, he says. “Google’s played a larger role than almost any company in bringing that empiricism into the mainstream of the field,” he says. “Because of machine learning and operation at scale you can do things that are vastly different. You don’t want to separate researchers from data.”

It’s hard to say how long Google will be able to count on luring leading researchers, given the flush times for competing Silicon Valley startups. “We’re back to a time when there are a lot of startups out there exploring new ground,” says MacCormack, and if competitors can amass more interesting data, they may be able to leach away Google’s research mojo.

This is more than a theoretical threat: last year Facebook hired away a leading deep learning researcher from Google to found its own research group in the area. However, Google’s piles of cash and its willingness to spend offer protection for the near future. “The fortunate thing for Google is they have this incredible cash-generating franchise that can fund a lot of experimentation and research,” says MacCormack.

Keep Reading

Most Popular

It’s time to retire the term “user”

The proliferation of AI means we need a new word.

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.

Sam Altman says helpful agents are poised to become AI’s killer function

Open AI’s CEO says we won’t need new hardware or lots more training data to get there.

An AI startup made a hyperrealistic deepfake of me that’s so good it’s scary

Synthesia's new technology is impressive but raises big questions about a world where we increasingly can’t tell what’s real.

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 with a list of newsletters you’d like to receive.