Skip to Content

Microsoft: Getting from "R" to "D"

The software giant’s Beijing lab is spearheading a new way to turn research into products.

Can a firm have too much innovation to handle?

It happens. And that problem was the dominant issue for a small group of executives from Microsoft’s Beijing research lab when they held a retreat in November 2002 at a spa in Zhuhai, not far from China’s border with Macao. Relaxing in the outdoor hot springs, the group brainstormed a promising solution to a fundamental problem facing Microsoft and many other high-tech companies: how to move more innovations more swiftly and effectively from research to development to market. Their idea: a new type of organization designed to bridge the gap between “R” and “D” and in the process overcome many of the product development bottlenecks and geographic and cultural differences that impede today’s global corporations.

Microsoft’s Advanced Technology Center (ATC) opened in November 2003 with 20 employees and a couple of projects. By late last year, after receiving more than 30,000 résumés from around China and sparking keen demand among Microsoft’s business divisions, it had around 100 employees, with some 17 major projects and scores of minor ones on its books; this year, the ATC is set to double in size. In the next few years, the center expects to be the key technology transfer point for a host of new products, from Web-search technologies to mobile applications and home entertainment systems. On the strength of these innovations, Hongjiang Zhang, the center’s charter director, hopes to provide a powerful alternative to Microsoft’s traditional strategy of creating products in the U.S., spiraling into Europe, and then adapting them for the Chinese market. “China is still emerging, but China is no longer just a follower,” he says. “They are starting to lead.”

So why not complement the current strategy by growing and testing new products in China and then introducing them in the United States?

Necessity: Mother of Development
ATC takes up half of one floor of a six-floor office building in the Haidian district of northwest Beijing – the same edifice occupied by Microsoft Research Asia, one of six research labs worldwide that Microsoft operates. Enhancing the transfer of technology between the 170-strong research facility and product development groups at Microsoft’s headquarters in Redmond, WA, is the whole reason the center exists, and the close proximity to the lab makes handoffs easier.

Zhang, who worked on technology transfer at Hewlett-Packard Labs before being recruited as a charter researcher at the Beijing lab, says the issues spurring the ATC’s creation are common to all research and development–driven organizations. After researchers hand off an invention or a new piece of code to product developers, a lot of refinement and testing is needed to get it ready for commercial release, and the product developers aren’t always able to do it. They often have their hands full with more pressing jobs – say, upgrading conventional features or improving security. And even a great invention might arrive at the wrong point in the product development cycle, where it’s difficult to fit into the next release. “Which means that for the product group to form a team and take that risk [of developing the invention] may be too big a risk and too big a distraction,” says Zhang.

His sentiment is echoed by Dennis Adler, general manager of business development for ATC. Based in Redmond, he is the liaison between the center and Microsoft’s product groups, helping the business divisions understand what advances are emerging from Beijing and the ATC members understand development schedules and constraints. Many times, he says, product groups would have loved to incorporate some cool feature or technology created in one of Microsoft’s research labs but just didn’t have the staff to devote to it. “There are tech transfers that haven’t happened because of it,” he says. “So ATC was set up as a way to help close that gap. It’s all about lowering the impedance.”

Microsoft officials say they know of nothing else like the Advanced Tech­nology Center either inside their company or outside it. “We just invented it,” says Adler. At least in Beijing, an organization like the center wasn’t even thinkable until a few years ago, and its creation shows the benefits of constantly adapting operations as conditions change.

Given that it typically takes several years for research projects to yield anything that can be commercialized, moving innovations to development was hardly a top priority when the Beijing lab was formed in 1998. Early on, though, a group was set up to help researchers build demos for showing concept technologies to their Redmond research colleagues and to the business divisions – with two engineers or programmers assigned to each of about a dozen projects. But as the lab grew, and its projects became serious contenders for product development, it became clear that while one project might require only two engineers or programmers, another might need five or more to handle the intensive testing and predevelopment refinement needed to verify that a one-of-a-kind research prototype could be mass produced.

The initial solution, in late 2001, was to form a centralized, more flexible organization dedicated to technology transfer. Rather than being assigned to specific lab projects, engineers in the new Tech­nology Transfer Group (TTG) went where they were needed – in the numbers needed. The new group marked a big step forward, says Zhang. But as the lab matured, more of its research teams produced more promising technology. “Now the scale is bigger,” Zhang says. So something bigger than the TTG was needed.

In addition to Hongjiang Zhang, the group that assembled at the hot-springs retreat in late 2002 included Ya-Qin Zhang (no relation), who headed the Beijing research lab from mid-2000 until early 2004, before becoming vice president of Microsoft’s Mobile and Embedded Devices division. They were joined by Harry Shum, then number two at the lab and now its director, and a handful of other key personnel.

Central to their thinking was China’s status as more than just an emerging market. Thanks to places like the Beijing lab, the country was making its mark in the creation of information technology, and the spa-soakers figured that a different type of organization, larger and more systematically woven into Microsoft’s product-development cycles than the Technology Transfer Group, might help the company take advantage of that creativity far more effectively. “That’s how the ATC concept started,” says Ya-Qin Zhang, who came up with the initial idea. “It’s an incubation center for new technology” – technology made in China.

Ballmer’s Key Recommendation
A few days later, Ya-Qin Zhang formalized the idea in a proposal to Microsoft’s senior vice president Rick Rashid, who oversees Microsoft Research’s worldwide operations. The plan focused on four prime goals: amplifying technology transfer by moving more products to development more efficiently; tapping Chinese talent in new ways; preparing for an emerging Chinese market hungry for cool new things; and incubating technology created in China in order to spread it to the rest of the world.

Rashid embraced the concept and asked for a detailed proposal. He then took the plan to Bill Gates, laying out the initial projects the center would take on and requesting a 50 percent increase in Beijing lab staff. The Microsoft chairman gave the plan the green light.

The last person to convince was CEO Steve Ballmer, who had to approve new personnel additions of this magnitude. Ballmer surprised Rashid by asking whether the 50 percent staff increase was adequate. His point was that by the time an innovation made it to the new Advanced Technology Center, it should be pretty much assured of becoming part of a commercial product. Therefore, unlike the research division, whose projects usually face a less certain future, ATC should have no rigid size limit. Ballmer told Rashid, “I am going to eliminate the ceiling.” Moreover, he noted, since the product groups would be the main beneficiaries of the new center, they should pay for the predevelopment effort.

Hongjiang Zhang says that Ballmer is “very smart” and that his funding idea proved particularly astute. When product groups pony up for an ATC project, he notes, they have a bigger stake in making it work, increasing the odds of success.

The Advanced Technology Center’s opening in November 2003 was timed to coincide with the five-year anniversary of Microsoft Research Asia. To help handle the flood of résumés that poured in, Zhang began holding written exams in 11 cities around China, with two sessions in late 2003, one in February 2004, and two in October and November 2004. Late last year, the center also began using one of the Beijing lab’s prototype technologies, a résumé screener that categorizes candidates by test scores and other parameters and gives priority to graduates of top computer science and engineering programs.

Zhang and Beijing lab managing director Harry Shum almost glow with excitement when they talk about the talent that those résumés represented. They wasted no time in putting their new hires to work. The first product to go to market was a video-editing technology that can easily summarize sports and news highlights, compressing an hour of video into five minutes. The software is now standard in Movie Maker 2.1, part of Windows XP. The Windows XP Media Center Edition 2005, a version of the XP operating system designed to make it easy to manage home entertainment from a PC, also features technologies created by the Beijing research lab and further developed by ATC. These include automatic image processing for digital photos and a function that automatically locates the center of attention in a photograph and highlights the area most likely to be cropped.

At the beginning of 2005, Microsoft was closemouthed about projects now in the ATC pipeline – projects it expects to yield commercial products within the next two years. But Zhang says six of Microsoft’s business divisions currently fund projects inside ATC.

One big push, he hints, will be in search technology. Microsoft researchers have been working in search for a decade and, says Zhang, the Beijing lab has been at it for five years. It’s only natural to expect that some of that work is nearing commercialization, presumably for Microsoft’s MSN Internet portal business and the Windows Client group, which is respon­sible for the company’s operating systems.

The Natural Interaction and Services division now run by Kai-Fu Lee, who founded the Beijing lab in 1998, is also sponsoring several projects. But ATC’s biggest customer is the Mobile and Embedded Devices division. “We’re putting a lot of focus on mobility,” says ATC assistant director Eric Chang. “China is the number one market with the handset, and the penetration rate is really low, so I think it will stay number one for a long time.” Chang says the center is building a team to work more closely with handset operators and manufacturers to more quickly create new technologies and get them into products.

Heading Microsoft’s mobile business, of course, is Ya-Qin Zhang, who conceived the Advanced Technology Center notion in the first place and is intimately familiar with the work being done in Beijing. Zhang couldn’t be happier with his creation. “ATC is a super success,” he says. “ATC now in Redmond is a star organization. Every product group wants to work with ATC.” Although the center is focused on Beijing lab creations, Microsoft officials say it could eventually handle innovations from the company’s other research groups in Cambridge, England; Bangalore, India; Palo Alto and San Francisco, CA; and even Redmond.

Future Concerns
It hasn’t all been smooth sailing. Hong­jiang Zhang says the biggest challenge so far has been helping the rush of new hires, many fresh from Chinese universities, fit into the Microsoft culture and development process. An English writer has been hired to proofread important e-mails and documents before they are sent. However, Zhang says, “We quickly realized the communication barrier is beyond just English. It’s really about culture differences. How do you follow up? How do you work with a team that is 16 hours away and follow through on all the deliverables? Or how do you simply say no clear and loud in technical or project discussions and learn not to overcommit?” To help with such issues, the center has created programs to train future managers, and senior managers say they work diligently to share their experiences with new employees.

Dennis Adler, the ATC liaison in Redmond, visits Beijing several times a year and hosts ATC visitors monthly. In addition, he relies on an array of communications tools – e-mail, voice over Internet, instant messaging, videoconferencing – to overcome the time and distance gap between his Beijing colleagues and him. “You just have to work a lot harder to keep the communication up,” says Adler.

And potentially rougher waters lie ahead. Managing the center’s growth will be difficult as it steams past the 200-­employee mark sometime this year. “Growing is easy, but to sustain it is hard,” says Bin Lin, ATC’s director of engineering. As more and more employees come on board, it becomes harder to maintain quality and meet deadlines, Lin says, and center leaders risk becoming so preoccupied with managing the here and now that they miss new opportunities.

Another problem facing organizations like ATC is that after a year or so of learning the ropes, their engineers might be lured away by Chinese firms, says Henry Chesbrough of the Center for Open Innovation at the University of California, Berkeley’s Haas School of Business. Because they are not doing real research, the engineers lack the benefits of a research environment that might entice them to stay, such as freedom from the tight deadlines of product development schedules, Chesbrough says. And because their job trains them to solve product development problems, they have skills that are enormously attractive to other companies.

Hongjiang Zhang is well aware of the potential pitfalls. But he allows himself a little time to bask in the center’s early success. Zhang says he was especially touched when Senior Vice President Rashid told him, “Without ATC, a lot of the things Microsoft should do and wanted to do will never happen.”

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.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

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.