The 2006 World Cup Final was among the most-watched sporting events in history, as an estimated 715 million viewers tuned in to see Italy knock off France, 5-3, in a penalty kick shootout.

The 2010 Cup Final, a 1-0 Spain win over the Netherlands, drew even more viewers—and ESPN, which owned the broadcast rights to this year’s tournament, seemed intent on learning as much as possible about the viewing habits of each and every one of them.

This spring ESPN announced the creation of ESPN XP, a new multi-partner research initiative through which the network intended to study consumer behavior tied to some of the biggest events in sports—the 2010 World Cup being the first. To study television viewership of the tournament, ESPN contracted with Nielsen Co., and to look at branding issues, it brought in marketing research firm the Keller Fay Group. But when it came to Internet and mobile technologies, “The Worldwide Leader in Sports” turned to a Wharton research center that may well be termed “The Worldwide Leader in Interactive Media.”

As one of the major components of ESPN XP, ESPN turned all data gathered about its mobile and Internet viewers during the first round of the World Cup over to the Wharton Interactive Media Initiative (WIMI), which moved quickly to analyze that data and then report back to the network with a prediction of how (and how many of) those same viewers figured to consume the later rounds of the tournament.

“This is obviously crucial information for them,” explains WIMI co-director Eric Bradlow, Wharton’s K.P. Chao Professor. “This is how they make money—by selling to advertisers.”

Glenn Enoch, Vice President of Integrated Media Research for ESPN, says the network is unique in that its brands cross basically every single media platform—television (through ESPN, ESPN2, ESPN U. and others), radio (via ESPN radio and a family of podcasts), Internet (through ESPN.com), mobile (through ESPN Mobile) and print (through ESPN The Magazine). As such, ESPN execs have long been interested in figuring out whether its television viewers, for example, are also heavy consumers of its other products.

The problem, he says, is the network hasn’t been able to find a dataset that can give them those answers.

“When we started looking at cross-media data, we only had information about each platform as silos,” Enoch explains. “What we lacked was the ability to see how sports fans were navigating from platform to platform. We want to understand and encourage that behavior. In other words, we want viewers of the television network to also be multi-platform users. It’s important to be able to tell advertisers on each of these platforms, ‘Here’s why TV works, here’s why radio works, and here’s why being a multi-platform advertiser somehow adds up to being more than just the sum of its parts.’”

The World Cup, and the partnership with WIMI, presented ESPN with its first opportunity to be able to do just that.

Peter Fader, Wharton’s Frances and Pei-Yuan Chia Professor and Bradlow’s fellow co-director, notes that while other large media companies have launched similar initiatives in the past, the ESPN-WIMI partnership is unique in that its goals are proactive. Rather than gathering data about a past sporting event, Fader says, WIMI was preparing forecasts for a future event. The unique structure of the World Cup—which takes about a month to complete—helped make that aim possible.

“NBC has looked back at the Olympics, trying to figure out how many people watched on television or via mobile or on cable,” Fader says. “But NBC was looking backward there, saying ‘Here’s who watched the Olympics on our various platforms.’ We wondered, ‘Wouldn’t it be better to say that before [the event] happens, so you could plan accordingly and inform your advertisers about it?’ We told ESPN, ‘Let’s do this proactively.’”

The 2010 World Cup wrapped up on July 11. But the conclusion of the tournament isn’t likely to be the end of ESPN XP. Already, ESPN has said that it will expand the initiative to include its coverage of both professional and college football. And Enoch says he’d like to see cross-media research fully integrated into ESPN’s daily research work by 2012.

“Instead of just being a special project, I’d like cross-media research to be something we can do every day—something that is part of our regular research initiatives,” Enoch says. “We’re just getting started with the World Cup.”

Bradlow and Fader, for their part, hoped the ESPN partnership—and the results of their World Cup work—will show other media companies how they stand to benefit from WIMI’s methods as well.

“At the end of the day, we believe this would be a success if we can predict the last three weeks of the tournament based on the first two weeks,” Bradlow explains. “If we can do that, we’ll [be able] to go around to other businesses and media companies and tell them, ‘We did this for the World Cup, and this is what we can do for your site.’ It would be proof of concept that academic forecasting algorithms have practical value and can help answer real business questions.” —T.H.