Services Case Studies Industries News Events Company Resources1 Insights  
newEvents gif newsEvents gif


News Coverage
Press Releases
Current Events
Past Events
  « back to news index

 

 

THE WASHINGTON TIMES/UNITED PRESS INTERNATIONAL | BLOG SOFTWARE TRACKS PREFERENCES

 

June 8, 2005

by Dan Whipple

 

The skyrocketing popularity of personal Web logs is providing new insights into the buying habits, political interests, and the social and cultural attitudes of important segments of the U.S. marketplace.

Umbria Inc. of Boulder has developed a Web crawler that monitors hundreds of thousands of blogs, turning the information from them, via natural language algorithms, into marketing data that is potentially much more reliable than traditional tools such as focus groups.

 

Umbria is tracking the market opinions of Generation Y -- the uncreative label for the youth of the nation currently between the ages of 10 and 27.

"We can parse the speech in these blogs, break it down by nouns verbs, adjectives and phrases to derive meaning and understanding about the speech and the speaker," David Howlett, Umbria's product management vice president, told United press International. "We use machine-learning algorithms to show who the speaker is and their characteristics."

The technology allows the company to determine both the speaker's gender and age range. It also tracks favorable and unfavorable product mentions to let companies know how their image is regarded among this important target age group.

 

For instance, in a recent release of data the company found that Starbucks, despite spending little on advertising, is the most mentioned non-alcoholic drink in the Gen Y blogosphere, ahead of much more visible advertisers Pepsi and Coke.

 

McDonald's is the fast-food brand mentioned most often, though not always as favorably as Wendy's and Taco Bell.

 

Umbria's Web crawler also confirmed what most baby boomers have long suspected: Their daughters' favorite celebrities are people the older generation barely recognizes, such as Britney Spears, Paris Hilton, Lindsay Lohan, Hilary Duff and Jessica Simpson.

 

Curiously, the company found that Gen-Y females talk more about beer than males do, with Guinness the most mentioned brand.

 

"What we've found interesting in tracking the blogosphere," Howlett said, "is that in Gen Y, the females were really the instigators in the use of Web logs. A little over a year ago, the majority of bloggers were women."

The blogosphere also has been quickly invaded by older folks and by males, so "Generation X and boomer males are starting to take over the preponderance of Web logs," he said.

 

Umbria is a privately held company with $6.7 million in capitalization. Current clients include Electronic Arts, Sprint and Logitech.

This mining of Web logs for data is not limited to private companies. Researchers at the University of Buffalo's School of Informatics also are studying how blogs reflect politics, opinion and attitudes.

 

"It is our contention that the totality of content across millions of Web logs vividly and objectively depicts the social landscape and ideology at certain points of time and space," Alexander Halavais, an assistant professor of communications, said in a statement.

 

Umbria uses a complex statistical-verification process to cross-validate whether the people monitored reside in the target group and whether the information gleaned is accurate. The results are not empirically verified, however -- that is, no one calls or visits the bloggers to see if they really are, say, a 14-year-old female in Altadena, Calif.

 

Nevertheless, Halavais' work indicates the blogs should be highly reliable sources of marketing information -- even a bit upscale. Halavais has found that blog distribution tracks population density fairly well -- more on the coasts, fewer in the middle. There are more blogs in suburbs and in high-tech hotspots.

He also concluded the densest concentrations of bloggers are associated with areas of high income, often more than $100,000 a year, on average.

"Existing methods of content analysis permit the codification of large amounts of blog text with its rich, unsolicited source material and easy accessibility," Halavais said, "and such codified, easily accessed material is a valuable resource for both macro and micro social-science research."

And for finding out whether kids prefer McDonald's or Wendy's.


 





 

 


McGraw-Hill Companies
  © 2008 J.D. Power and Associates, The McGraw-Hill Companies, Inc. All rights reserved. Site Map | Privacy Policy