Media organizations routinely make crucial editorial and business decisions based on Web usage and traffic. Yet precisely measuring the size of Internet audiences can be difficult, and it is often fraught with conflicting data. Interpreting such fractured online metrics affects how news organizations understand their audiences, evaluate their competitors and sell advertising.

A 2010 study by Columbia University’s Tow Center for Digital Journalism, “Confusion Online: Faulty Metrics and the Future of Digital Journalism,” explores how different measurement companies and services arrive at entirely different findings in terms of Web traffic. The study looks at how measurement tools can produce statistical outcomes that can vary by more than 100% and examines the steps necessary for a more navigable digital media future.

The study’s findings include:

• Leading firms such as Nielsen NetRatings, comScore, and Hitwise, and services such as Google Analytics and Omniture, are competing in the same market space and often produce greatly diverging results; no single “currency” of measurement has yet emerged for media companies.
• Results vary widely because data is measured at different tiers by different companies: some measure traffic at the publisher server level; some measure traffic data at the level of ISPs, or service providers such as Verizon or Time Warner; and still others use tracking software to follow designated panels of Internet users, whose patterns are then extrapolated.
• Examples of divergent Web traffic numbers are: In October 2009, Nielsen NetRatings put The Daily Beast’s audience at 1 million, but comScore said the site had 2.2 million visitors; during one month in 2010, Google Analytics said Talking Points Memo had 1.8 million unique visitors, but comScore said the site only typically gets around 300,000 a month.
• News outlets such as the Miami Herald, Washington Post and Wall Street Journal use multiple, often-conflicting, measurement services or firms. Some digital managers report that this environment makes decision-making difficult.
• The crucial measurement of “unique visitors” remains in dispute; influential organizations such as the Web Analytics Association and Interactive Advertising Bureau have traditionally had different definitions.

The authors conclude that there remains some possibility of a more definitive measurement emerging and “hybrid” combinations of measurements offer some promise. What media organizations most need, the authors state, is more detailed demographic and behavioral information about visitors to their sites.

Tags: technology

By John Wihbey | April 12, 2011