Market researchers comScore Networks Inc. and Nielsen//NetRatings Inc. are battling for the best Hispanic Internet panel, but one is taking the fight a step further.
Going on the offensive, comScore claims its panel to measure Hispanic behavior online is superior to NetRatings' five-year deal announced April 2 with media owner Univision Online Inc. An April 4 letter sent to clients of comScore Media Metrix Hispanic Services was clear.
“Their announcement claimed that the service would be the first of its kind. But as you know, the reality is markedly different: with your support, comScore Media Metrix had already introduced an Hispanic measurement service in October 2002,” Richard Israel, vice president of Hispanic marketing solutions at comScore, said in the letter.
The letter from the McLean, VA, company stresses comScore's commitment to an unbiased, accurate understanding of the online Hispanic population. It adds that comScore “cannot allow misinformation or biased approaches to turn back the clock.
“We have heard, in no uncertain manner, that many of you are concerned about the implications of one publisher sponsoring the creation of an 'industry' information source,” Israel said. “In contrast, by building a broadly endorsed service, we believe we have built an industry-accepted standard that best serves the interests of the entire of the [sic] Hispanic marketing community.”
So what is the cause of such sniping?
NetRatings has partnered with Univision's univision.com, a leading Spanish-language Web site in the United States and a sibling to the TV channel of the same name. The aim is to build an Internet audience measurement panel to improve tracking and analysis of online Hispanic behavior nationwide.
The panel will be based on random digit dial methodology, with NetRatings deploying its strategy to a representative panel of 3,600 Hispanic individuals. The sample will be included in the current Nielsen//NetRatings panel that represents the U.S. online universe.
NetRatings said the panel will offer demographic information on Hispanic Internet users. It will serve media buyers, planners, publishers and marketers targeting online U.S. Hispanics, the nation's fastest-growing ethnic segment.
Partnership with Univision, NetRatings' first client for this panel, will help. Nielsen Media Research first struck a relationship with that company in 1992 to benchmark U.S. Hispanic television ratings.
The panel will recruit between now and midsummer for October data to be released the following month. The syndicated research will be sold to any interested company.
“This is supposed to be a syndicated panel, and this will be the only panel to recruit people over the phone in both English and Spanish, and no one else is doing that,” said Jed Meyer, senior vice president of business development at NetRatings in New York.
ComScore's panel, announced Dec. 17, has partnered with the Association of Hispanic Advertising Agencies for a representative online panel of 50,000 U.S. Hispanics. Members of the association get a special introductory subscription as part of this cooperation.
The comScore database will offer granularity into the behavior of this ethnic segment at myriad sites, including AOL. It also will track online purchase activity for allowing advertisers to gauge a Hispanic site's audience based on its actual buying power.
To enable such monitoring, comScore panelists have downloaded an applet into their browser for reporting online activity.
Cities with Hispanic concentrations like Los Angeles, New York, Miami, Houston and Chicago will provide fertile grounds for data, comScore claims. Data on Puerto Rico will be offered as a supplement to the U.S. service.
But the biggest advantage of comScore's panel may be its running start. It has charter clients like AOL, The Bravo Group, Media 8 Digital Marketing, Terra Networks, La Opinion, Communita, Yahoo en Espanol and Bromley/Tapestry.
And then there are claims.
ComScore said its panel has a large representative sample that includes random digit dial recruitment methodologies. The panel has been selected and weighted to represent five language segments, country of origin and geographic location of U.S. Hispanics.
But it is the Univision connection with NetRatings that most rankles comScore.
According to comScore, the Univision panel will measure only home use. ComScore will measure that as well as at-work and college/university use. Its sample includes 3,000 to 5,000 Hispanics at each of these locations.
Next, the Univision sample for Spanish-language-preferred Hispanics may include only a few hundred active users in each segment, comScore claims. This number will drop when measuring specific sites, giving way to small samples beyond the largest Web sites.
Finally, comScore takes issue with size. The Univision panel is only 1/14th the size of its own panel of 50,000 U.S. Hispanics. This larger size represents more accurately behavior in each language segment. However, comScore's estimated monthly active online sample is only 32,500. NetRatings' active monthly sample is to be 2,520.
Of course, comScore has only 5,000 random digit dial panelists on its panel, which is more than NetRatings' entire panel size. Online recruitments account for the rest.
“We use those RDD panelists to calibrate data from the balance of the panel,” said Dan Hess, vice president at comScore. “The virtues of this approach, which we created with our global sample more than three years ago, have been well-established.
“Nielsen itself has borrowed from our approach with the announcement of its 'megapanel,' supposedly to be introduced later this year,” Hess said. “Why this isn't being used in the Univision Hispanic panel I can only speculate.”
NetRatings' Meyer did not want to get into a “he-said, she-said” match. But he noted that it costs roughly $1,000 to recruit someone in Spanish over the telephone versus the Internet. And telephone recruitment is superior to online, he said.
“We're recruiting everyone over the phone in English and Spanish,” Meyer said. “The difficulty is to get representation. It's universally known that opt-in online research is not as representative as telephone and print recruitment.”
