By: Jennifer Saba
Anyone familiar with online rankings knows that, depending on the measurement service, the data tends to swing wildly. It’s a problem not just with newspaper Web sites but also with the larger online world. The Interactive Advertising Bureau (IAB) has been trying to hammer out a solution to address the problem where advertisers can effectively glean where the trends are going.
As the latest example, big differences appear in the August traffic data for top newspaper Web sites from comScore and Nielsen//NetRatings — never mind the internal data from the companies.
Yesterday E&P covered a somewhat troubling report from Wachovia that tracked monthly uniques and page views — using comScore — for newspaper companies in the research firm’s coverage universe. For the most part, the data proved that newspaper Web site traffic slowed in August.
Today, E&P obtained the August traffic numbers for the top 30 Web sites from Nielsen//NetRatings (owned by E&P’s parent company) to see if the slowdown was a fluke or the start of a more distrubing development. This is not to pit comScore against Nielsen or to say that the data from one is more meaningful than the other. Both comScore and Nielsen base their traffic and page views on panel-based surveys.
Just looking at the percent change from in the August unique visitors for about seven newspapers shows that it’s a mixed bag.
According to comScore, uniques at the NYTimes.com fell 11.5% in August year-over-year; Nielsen says they were up 6%.
Both outfits show that unique audience grew in August compared to August 2006 at The Wall Street Journal online. Just how much is up for debate: comScore said traffic at the site advanced 15.9%. But Nielsen recorded an advance of 54%.
USA Today’s unique audience dropped in August, revealed both comScore and Nielsen, down 13.6% and 5%, respectively. But USA Today Senior Vice President of Marketing Susan Lavington counters that Gannett’s national paper “had a fabulous August” with internal logs showing traffic up 3% in August compared to July.
USA Today switched the software used to track their numbers this year so they don’t have yearly comparisons yet.
Lavington says that Nielsen’s numbers never line up with their internal logs but she offers an explanation as to why both Nielsen and comScore are showing decreases in August traffic. USAToday.com dropped a distribution deal the paper had with the ISPs Juno and NetZero — two providers that tend to cater to home users. Nielsen’s and comScore’s data lean heavily on home users to make up their panels since it’s difficult to get companies and business to allow the measurement firms to track activity on work computers.
Christian Hendricks, McClatchy’s vice president of Interactive Media, points out that flaws will appear no matter who is holding the calipers and that data should not be considered set in stone. “The numbers are directional and it’s just a conversation starter,” he says.
According to comScore, McClatchy’s traffic declined 25.5%. Hendricks said that McClatchy’s internal logs — tracked by Omniture — shows that uniques were actually up 18% year-over-year.
He attributes the spike to an aggressive marketing campaign rolled out companywide after executives realized that in 2006, traffic was slowing.
Randy Bennett, vice president of audience and new business development at the Newspaper Association of America, notes that in general the growth rate of newspaper Web sites is decelerating but that there is still an upward trend. “As any medium matures, clearly the growth rate is going to start slowing,” he said. “[Newspapers] can still attract people but maybe not a rate of 20% a year maybe more like 10% or 15%.”
News is cyclical too. Bennett thinks that in 2008, newspaper Web sites could see a surge in unique visitors thanks to the Olympics and political year.