The world of pro sports long ago recognized how “Moneyball” tactics can transform a franchise -- like baseball's Chicago Cubs rising from lovable losers to hosts of a National League Championship game Wednesday -- but they're a relatively new trend in the business world. A growing number of companies and executives are using nontraditional methods of data analysis to cut costs, improve employee performance and attract new customers.
“Moneyball,” known in baseball circles as “sabermetrics,” centers on the use of advanced stats to more accurately gauge on-field performance and identify undervalued players. The concept exploded in popularity with the 2003 release of Michael Lewis’ book, which documented how Oakland Athletics general manager Billy Beane used statistics like on-base percentage, rather than more traditional measures of success like home runs, to turn his underfunded team into a perennial contender. Since then, advanced analytics have caught on in the NBA and other major American sports leagues, as well as in international soccer and tennis.
With the uncertainty of today's financial landscape, statistics that can help companies eliminate waste or identify new customers are hugely valuable. Employees at online retailer Amazon are parsing data sets such as customer "click rates" and email-open rates to improve their marketing and decide which products to keep in stock, with $88 billion in revenue to show for its efforts in 2014. Organizations like Teach For America apply advanced metrics to target the best job candidates, while medical professionals use real-time biodata to improve how they monitor patients. The possibilities, quite literally, are endless, and companies are still learning how to sift through mountains of information.
“We don’t have a data problem anymore. We have a ‘too much data’ problem,” said Fred Feinberg, a professor of marketing and statistics at the University of Michigan’s Ross School of Business. “Data is not information. The key is, when information is coming in so fast you can’t even store it for more than a few days, you have to decide what to retain.”
Expansion In Sports
Data of every kind is more abundant than ever, and executives can apply their research in any number of different ways. For Beane, whose Athletics had the third-lowest payroll in baseball in 2002, that meant finding player like catcher-turned-first baseman Scott Hatteberg, who could produce at a high rate on the field at a bargain salary.
In the National Basketball Association, Houston Rockets general manager Daryl Morey spearheaded a statistics-based approach that has changed the way teams play the sport. With the backing of computing systems that can quickly parse large data sets, teams have eschewed the traditional two-point jump shot in favor of three-point shots and layups, which are seen as more efficient. Players who can perform within that new system have seen their earning potential increase exponentially.
“Moneyball” is starting to catch on internationally, too. Tennis players use big data to streamline their on-court movements and identify weaknesses in their games that can’t be seen from basic statistics like unforced errors. Sabermetrics are still in their infancy in soccer, but efforts are underway to integrate them.
It’s taken just a few years for companies to realize data analysis can be a versatile, powerful tool, with all sorts of potential applications. The “Moneyball” concept is now essential to the hiring process, as companies seek to identify strong job candidates, maximize employee work output and retain workers who prove to be valuable assets.
Companies are parsing job boards and resumes for attributes found most often in their best employees. For example, an international accounting firm could determine that applicants who went to a certain school or possessed certain foreign-language skills tend to be particularly productive, just as a baseball team would target players with a high on-base percentage. If a statistical analysis shows employees are leaving a company because a competitor is offering more money, that company can restructure compensation and perks to keep workers happy.
Investment banking firms Goldman Sachs and Credit Suisse, tech front-runners like Google and Microsoft and nonprofits like Teach For America are just a few organizations that have embraced this sort of employee analysis, said Cade Massey, professor of operations, information and decisions at the University of Pennsylvania’s Wharton Business School and host of “Wharton Moneyball” on Sirius XM radio.
“Many organizations are trying to better understand what really predicts employee success. People are more and more aware that the traditional resume-and-interview approach leaves a lot to be desired,” Massey said.
As consumers increasingly turn to Internet outlets to do their shopping, advanced metrics help online retailers eliminate waste from the supply chain and reduce expenses. By analyzing product reviews, search histories and previous purchases, companies like Amazon can tailor the marketplace to suit an individual customer’s preferences.
Most online retailers have developed algorithms that provide “recommended products” after checkout, in hopes of luring customers into spending more money. But Amazon employees also make extensive manual use of stats like "click rate" and "opt-out" rate, or how often customers unsubscribe from email lists. If a particular customer buys a lot of books and video games, Amazon determines which product purchases will yield more revenue and alter its email lists to match, an employee told Fortune in 2012.
“Amazon is a company that is built fundamentally on data-driven decision-making. It’s because of the trove of information that they have accumulated about consumer behavior based on consumer purchases on that platform,” said Ben Ryan Shields, a data analytics expert and lecturer at the Massachusetts Institute of Technology’s Sloan Management school.
This sort of data-driven analysis can also yield massive dividends from a marketing and advertising perspective. Google, which has embraced advanced analytics from the start, can help potential partners identify how and when to advertise their products based on real-time search data. For a company that generates as much Web traffic as Google, the potential for targeted advertisements is staggering.
In one case, Google determined which regions of the country were most affected by the seasonal flu based on where residents were doing online searches for symptoms. The company identified outbreak sites even before the Centers for Disease Control did, said Feinberg, the University of Michigan professor. That information could be very valuable to a company selling over-the-counter flu medication.
“If you want to talk about clever data usage, they are absolutely number-one in the world,” said Feinberg. “They can find out minute-to-minute how people are interested in new products and what they’re looking for.”
The rise of advanced data analytics has also had a perceptible effect on the medical industry – particularly in terms of how doctors treat their patients. More and more healthcare providers are switching to electronic medical records or “EMRS,” which place a patient’s treatment history and notable traits in an easy-to-read digital format. It’s a development that “could be a massive revolution for the way the healthcare industry is run,” said Shields, the MIT lecturer.
Wearable technology, such as the FitBit, provide users and doctors with real-time health data on everything from heart rate to sleep patterns. Doctors can now monitor health statistics on a constant basis, rather than during an hours-long visit or a hospital stay. Andre Iguodala, the Golden State Warriors shooting guard and 2015 NBA Finals Most Valuable Player, credited wearable technology with helping team doctors keep players healthy and well-rested.
A Work In Progress
While new data analysis methods are revolutionizing the way many companies do business, it’s still a relatively new concept that occasionally presents flaws. The usefulness of statistical research is in the eye of the beholder, and plenty of executives are still figuring out how to turn troves of information into actionable strategy. Companies that are using data most effectively tend to conduct proactive marketing experiments, rather than simply staring at shopping data in the hope that a solution presents itself.
Data is also easy to misinterpret, which leads some companies to employ ineffective strategies or waste resources based on faulty logic. Analysts often mistake correlation and causation, said Joel Shapiro, executive director of the data analytics program at Northwestern University’s Kellogg School of Management in Illinois. For example, a clothing company could theorize that customers who received a lot of store emails were more likely to buy products, but those customers might only be on the email list because they bought something in the past.
As companies like Google and Amazon turn data-driven research into tangible cash, smaller firms might try to adopt the same tactics. But too much reliance on statistics might also lead business leaders to adjust their perception of the “Moneyball” approach.
“Too many people think that data analysis and analytics is a panacea for all their problems,” Shapiro said. “I think analytics, while very powerful, is a bit overhyped. People are going to soon realize that analytics can’t solve everything.”
In the meantime, every tech startup and investing juggernaut will try to use “Moneyball” to beat their competitors. Especially if the Cubs win the World Series.