HP Labs' Latest Experiment: Itself

As the computer giant cuts its number of major research projects, its labs director explains the attempt to marry the grant-writing focus of a university with the entrepreneurial approach of a venture capital firm
Prith Banerjee, director of HP's research labs Timothy Archibald

Until recently, Hewlett-Packard's (HPQ) research labs were one of the few pockets at the company untouched by the efficiency mantra of Chief Executive Mark Hurd, who has taken a sharp pencil to HP's operations. But that's about to change—in a big way. On May 1, HP's new labs director, Prith Banerjee, plans to unveil a list of 20 to 30 major projects HP Labs will pursue, a dramatic cut from the 150 or so currently on its docket.

To shake up the slow-moving research and development group and make its efforts more commercially relevant, Banerjee is overhauling HP's research slate. The labs' $150 million annual budget will remain the same, but he'll group the most promising related projects while dropping those with little shot at a profitable payoff. "Just because it's scientifically interesting won't do it," Banerjee says. "We need to create whole new business opportunities for HP." No layoffs are planned, but researchers whose projects get scuttled will be transferred to bigger teams. At a time when every dollar must go further, HP's approach could be a template for others looking to wring more out of R&D as the economy turns south.

In the past, HP Labs relied too much on personal relationships among its 600 scientists when giving projects a go-ahead, says Banerjee. Researchers were subject to few formal milestone reviews. Projects often dragged on, making it tough to staff new ones. "Things were a bit territorial," says Jaap Suermondt, who heads an HP data analysis lab. And coordinating scientists' work with headquarters could be a struggle. "Convincing the business units took almost as much energy as the idea itself," says Ajay Gupta, director of HP's India lab.

Now Banerjee, a former engineering professor and founder of two chip design companies, is trying to marry the grant-writing focus of a university with the entrepreneurial approach of a venture capital firm. Researchers will compete for money and manpower by pitching projects, complete with written business plans, to a central review board that will approve ideas and monitor progress. A new technology transfer office aims to speed promising projects into HP's product groups, and license those that it can't use itself. A novel "entrepreneur in residence" program with VC firm Foundation Capital, based in Menlo Park, Calif., will help plug HP into startups that could commercialize tech from its labs.

Under HP's revamped labs, hierarchies will be flatter, with the seven global outposts reporting directly to Banerjee. Projects will focus on five strategic areas: intensive data analysis, building clusters of machines that can tackle complex problems, transforming analog into digital content, mobile computing, and environmentally friendly machines. Accountability will also be stronger: Researchers who go for too long without a funded idea risk expulsion to more mundane product engineering groups. "If not enough progress is made, the plug is pulled," Banerjee says.

Such high stakes could jeopardize that most tricky of corporate balancing acts: How to grant scientists leeway to spot technical shifts without letting them beaver away for years on projects that don't deliver measurable results. If it goes well, Banerjee's bet could help churn out the technological breakthroughs HP will need as the growth in tech spending slows and as Hurd's operations cuts—he's already lowered costs significantly—run their course. "This is the Mark Hurd HP, and sometimes you have to disrupt something to get to the next level," says Suermondt. "It's getting harder to move the needle."

Return to Inside Innovation Table of Contents

Before it's here, it's on the Bloomberg Terminal.