DuPont Tries to Unclog a Pipeline

Can it move new products out of the lab faster?

Whether it was the creation of nylon in the 1930s, Lycra in the 1950s, or environmentally friendly herbicides in the 1980s, chemical giant DuPont (DD ) has a history of developing products that become big moneymakers. This is, after all, the company that got started because its founder figured out how to make a better batch of gunpowder back in 1802.

Yet in the past decade, DuPont's innovation machine seems to have shot mostly blanks. Other than a drug for high blood pressure developed by its now-divested pharmaceutical unit, few big sellers have emerged from the company's $1.2 billion-a-year research and development operations. DuPont simply diverted too many resources to projects aimed at boosting productivity in existing businesses rather than developing new projects. And a failure to read market demand in some key areas, particularly in the agricultural biotechnology field, led to missed opportunities and questionable investments. "It's been a huge amount of money expended and very little visible output," grouses Nicholas Redfield, equity analyst at Banc One Investment Advisors, which recently held about 1.4 million DuPont shares.

DuPont needs some of its old-time R&D firepower--quick. In 2001, it sold off its drug business when it realized it was too small to compete effectively against giants such as Merck & Co. (MRK ) and Pfizer Inc. (PFE ) DuPont wildly overpaid, analysts say, for a new seed business, while its ag biotech operations have yet to live up to expectations. Meanwhile, DuPont has said it will divest some big businesses, such as nylon and polyester, that total about 25% of sales but are low-margin commodities. Nylon, for instance, now yields estimated aftertax operating margins of only 2%.

Now DuPont chairman and CEO Charles O. Holliday Jr., who took the helm in 1998, hopes that a rejuvenated R&D operation can boost growth. He elevated a DuPont lifer, Thomas M. Connelly Jr., to head the operation in the fall of 2000. Connelly, who started as a researcher, most recently ran DuPont's big fluoroproducts operation. Holliday has promised that products introduced in the last five years will account for one-third of revenues by 2005, up from 24% now. That is critical if DuPont is to meet its long-term goals of 6% revenue growth and 10% profit growth. "We can't get there without it," Holliday says.

Connelly's plan is to deliver a steady stream of new products--a lot of singles and doubles--while hoping for a grand slam along the way. Among the products so far: a new fabric for use in hospital operating rooms, a new form of auto-paint coating, and a new generation of fuel-cell components. The company figures such projects can generate north of $30 million in annual sales on average, but it will need a lot of those hits to meet the targets Holliday has set. "I like what they are doing," says Sanford C. Bernstein analyst Graham Copley. "I'm just not convinced we'll get enough [new products]" to make a big difference.

This year, earnings from continuing operations, minus one-time items, should increase 18%, to $2.25 billion, on 5% higher sales of $27.4 billion, says Copley. But that's coming off a depressed base. Profits in recent years were hammered by high energy prices and a weak economy. Indeed, income from continuing operations in 2003 won't even match the $2.9 billion recorded in 1998. DuPont shares, at $44, are down 26% since Holliday took over in early 1998, vs. a 5.5% drop in the Standard & Poor's 500-stock index.

To rev up the product pipeline, DuPont is reallocating its research dollars. Throughout much of the 1990s, two-thirds of its R&D budget went to improve productivity at existing units. Only about one-third was spent on new products. "That was a short-term move to get earnings up," says Joseph P. Glas, who retired four years ago as vice-president for life sciences research. "It was a big mistake." The split is now about 50-50, and the goal is to spend about 65% on new products.

Connelly is also focusing DuPont's research more carefully. His team has identified 75 projects that have the highest revenue potential. One of the 75 is Suprel, a lightweight and puncture-resistant fabric that will be used in gowns for surgeons and nurses. Lori A. Gettelfinger, global marketing manager for medical fabrics, says DuPont gave her the resources to move Suprel from concept to marketed product in just three years. The product hits the market this quarter.

But DuPont needs more than speed. It must also, say its critics, learn to correctly read the markets it wants to serve. Back in the late 1980s, for example, DuPont was one of the first to come up with a machine for sequencing DNA. That could have been a huge business, but the company never managed to commercialize it. And after paying $9 billion for seed company Pioneer Hi-Bred International Inc. in the late 1990s, development of new genetically modified seeds ran up against a major consumer backlash, especially in Europe. That has stalled development of some transgenic seeds that grow more nutritious crops. "Their science is superb. But they misread the market," says Gene Pisasale, senior investment officer at Wilmington Trust Corp., another big shareholder.

No surprise, then, that Connelly is pushing DuPont to work more closely with its customers. Two years ago, one of its teams developed a more durable and eco-friendly clear spray coating for DaimlerChrysler's (DCX ) Dodge Durango. DuPont Product Manager David W. Wood says the auto maker needed a new coating to help meet stringent air-pollution requirements in Delaware, where the car is made. DaimlerChrysler was pleased, and now DuPont is talking to other carmakers about using it. Bernstein's Copley figures the coating could yield up to $100 million in annual sales.

That is a sign of progress. But Connelly needs to prove DuPont can regularly repeat that success. DuPont has long had the science. Now it must develop the commercial savvy to match.

By Amy Barrett in Wilmington, Del.

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