The Issue: Setting IBM's R&D Agenda

New director of research John Kelly had to refocus the company's R&D without alienating scientists. The solution? Collaboration and diversification

Watch the Video…

John Kelly knew he had some big shoes to fill when he took over as only the seventh-ever director of research at IBM (IBM) in the fall of 2007 (BusinessWeek, 2/28/08). With a team that includes the winners of five Nobel prizes and more than a dozen National Medals of Science & Technology, he was heading up a stellar group. Nonetheless, Kelly's major challenge was to balance IBM's accolade-studded history in R&D with a rapidly changing business environment.

Unlike many of its former research rivals such as Bell Labs (now owned by Alcatel-Lucent (ALU)), which have scaled back on research, IBM continues to make very large investments in the discipline—to the tune of $6 billion a year. But IBM is up against new competition, including companies such as Google (GOOG) and Microsoft (MSFT) that are investing heavily and gaining ground on IBM in the race for patents. (IBM has been the leading recipient of U.S. patents for 15 years running, but last year Microsoft landed at No. 6.) To keep the company's lead, Kelly realized the company had to be able to transform and reinvent itself without an increase in the research budget. And he couldn't do so without the support of his team of creative and opinionated researchers.

Kelly spent the fall of 2007 visiting the eight IBM research labs based in countries including China, India, and Israel, talking directly to the 3,200 researchers in a series of town-hall-style meetings and appointments. From his travels, Kelly garnered a few key insights. No. 1: IBM needed to become even more global. Second, its annual R&D budget of $6 billion was dispersed across all sorts of projects. Kelly decided the company should make fewer, bigger bets that might be risky, but could yield bigger breakthroughs.

Balancing Priorities

Kelly, a PhD in materials engineering himself, decided the company needed to realign its research agenda, balancing very long-term projects (up to 20 years) with short-term ones (a year or two) and very large projects with smaller ones. And he honed the company's approach, whittling away the number of focal points for his research teams. At the beginning of 2008, he placed four big bets. Over the next few years, IBM will channel more than $100 million into just four areas, from working on core nanotechnology to developing business processes for companies struggling with the massive amounts of data necessary to operate in the modern world.

Kelly also realized that globalizing research was not about building more large brick-and-mortar labs. He wanted more agile, in-market research so he invented "collaboratories," centers that draw on the existing resources of universities, companies, and countries to tackle certain industry-specific problems. For example, instead of tackling transportation issues in the U.S., where the infrastructure is in line with growth, IBM will focus on a country, city, and area where road traffic is a primary problem, such as Mumbai.

Getting His Team on Board

With his new agenda in place early in 2008, the next challenge Kelly faced was getting his highly talented, but hugely cynical, employees to buy into his plan. The proposed streamlining was going to mean significant change for many of the researchers—with a real risk that some might feel marginalized if their projects were not among the top four priority areas.

In January, he hosted two 2-hour meetings from IBM's lab in Yorktown Heights, N.Y., each held at a different time to accommodate the various time zones of external labs. All members of all of IBM's laboratories were connected via video link and were encouraged to ask questions and participate. It was a logistical nightmare to organize, but Kelly is convinced the open presentations helped him get the researchers on board as they realized their work was valued and important, and understood he was making the right bets for all of them.

McConnon is a staff editor for BusinessWeek in New York.