Ford taps new 'Whiz Kids' to help navigate road to success

Bryce G. Hoffman / The Detroit News

Dearborn --Ford Motor Co. has a secret weapon: a group of latter-day whiz kids who are using mathematical models to boost the amount of money it can command for its vehicles.

Led by Rose Peng, this team of 19 Ph.D.s from the fields of mathematics, physics, statistics, economics and engineering is helping the company set pricing, plan production and strike the right mix of options and features in its new cars and trucks. Other automakers haven't committed similarly qualified employees in the way Ford has.

The Ford team is credited with helping the automaker post a $1.9 billion net pricing gain in the first half of one of the worst years in automotive history. "Rose and her team have been key to that," said Ford sales and marketing chief Jim Farley. "They are unbelievable. It's very scientific. I've never seen anything like it in our industry."

But it is not the first time Ford has looked to mathematics for salvation. Six decades ago, a group of young Army Air Forces officers led by statistician Charles "Tex" Thornton famously paved the way for the Dearborn automaker's postwar resurgence and created modern corporate management while they were at it. Dubbed the "Whiz Kids," they were the first to apply mathematical models to corporate decision-making.

Now, Peng and her team are taking that science to a whole new level. "We forecast the future," she said.

The 'Silent Lamb'

Born in Beijing, Peng was trained as a mechanical engineer in China before coming to the United States in 1987. She began working as a database consultant for Ford 10 years later and was brought on as part of the newly formed Global Lifecycle Analytics Department, or GLAD team, in 2000. She soon earned the nickname "Silent Lamb" because she never spoke in meetings.

"I was one of the first group members, and I had so much to learn," she recalled. "Our group was new, and the head executive wanted to understand why there was a gap between Ford products and the best-in-class competitors when it came to residual value. We had no data, nothing to go on."

By 2005, they had figured out that the resale value of Ford's cars and trucks was being eroded by sales of poorly contented vehicles to rental agencies -- a critical problem for a company taking back a million leased vehicle each year. Their models showed that if Ford cut back on these low-margin sales, the value of its vehicles would go up dramatically.

And it did.

Soon, Ford executives were asking what other insights the GLAD team could provide. Peng and her colleagues soon began developing models for production planning, marketing spending and vehicle pricing.

"All of those decisions are enormously complex and interconnected," said Wallace Hopp, professor of manufacturing at the University of Michigan and an expert on mathematical modeling. "Historically, the way companies have managed them is by decoupling them. But that is not effective. Mathematical models can help companies understand the whole picture."

American executives have been dismissive of such analytical approaches. Other U.S. car companies have similar talent on staff, but Hopp said none has put them to work on concrete business problems the way Ford has. The same cannot be said of Japanese automakers, who were quick to appreciate the potential of this approach to corporate planning.

"The people who are able to use it effectively are going to beat your brains out -- and for years the Japanese have," he said.

But even the Japanese were sent reeling by the current economic crisis, which makes Ford's pricing gains all the more impressive.

Less art, more science

Peng and her team identified several ways Ford could raise net pricing -- the actual price consumers paid for a vehicle less any discounts or incentives -- and give customers more for their money at the same time. They developed mathematical models to help the company better match production to the actual demand for its cars and trucks; figure out how much money to spend on advertising and incentives; decide which features to include in each trim level; and determine how much to charge for the finished product.

"Before, it was more art than science. Now it is definitely more science-based," Peng said.

"Instead of just trying to sell the vehicles to push up volume, we are actually giving the customer higher value with their product and pulling back incentives. At the end of the day, they will have lower cost of ownership and higher residual value."

The results might seem counterintuitive, given the challenging economy. Ford has added features to some new models like the redesigned 2010 Ford F-150 pickup, and raised the price. With other models like the new Taurus, it has kept the base price the same but added expensive new trim levels such as the SHO edition, which costs over $12,000 more. And it has slashed its incentive spending even as its competitors were increasing theirs.

Though its sales fell along with the rest of the industry's, they did not fall as much. Ford's share of the U.S. market increased, as did the money it made off of each sale.

"Our strategy at Ford has been fundamentally different from our competitors this year," Farley said, adding that Peng and her colleagues deserve much of the credit, along with Ford's new models and improved quality.

Hopp says Ford's comparative success demonstrates how valuable this scientific approach can be, but he said one need not look far these days to find the limits of mathematical modeling. Faulty mathematical models are being blamed for precipitating last year's economic collapse.

"If you went to any investment bank last year, it was full of Russian physicists," he said, explaining that Wall Street became too dependent on mathematical models it did not understand -- models that suggested U.S. home prices would continue to appreciate indefinitely.

Ford also learned the limits of statistical analysis when the original Whiz Kids tried to reduce every corporate function -- including vehicle design -- to a mathematical equation that contributed to the company's bottom line. The result was vehicles like the infamous Edsel that only made sense on a spreadsheet.

Peng said she and her team avoid such pitfalls by testing their models with historical data and adjusting them in real time.

As for the Whiz Kids, they all held more senior positions at Ford than Peng and her colleagues, and wielded far more power in the company -- though Ford's top executives now listen closely whenever the "Silent Lamb" speaks.

"I don't think we can compare ourselves with them," she said. "It is enough of a reward for us when the company uses our models."

bhoffman@detnews.com (313) 222-2443