NetRatings Seeks to Reconcile Online Audience Measures

Nielsen//NetRatings has developed a new method of measuring Internet audiences, integrating data from the two most commonly used approaches: data from consumer panels and site-centric measurement technology. Different methods can produce very different statistics, as seen in the latest search engine usage data from competitors Hitwise and comScore.

The new metric, which has been trialled in Italy, will be launched in Australia next year, and is being tested in the US. It combines an online consumer research panel, NetView, with a web analytics platform, SiteCensus. NetView provides data on consumer demographics and behaviour, along with comparable market data for 11 countries, based on Internet surfing from home and work. SiteCensus offers web site owners a detailed insight into their own audience.

One problem of current measurement methods is caused by the increase in cookie deletion by Internet users. If a user deletes a cookie, the web site will not recognise them as a repeat visitor on their next visit. This can lead to an overestimation of audience figures and an underestimation of visitor frequency. By linking a site-centric solution with a panel, Nielsen//NetRatings hopes its new system can provide a more accurate measure. Manish Bhatia, Senior VP of Product Development and Measurement Science, says the new Data Integration model takes the best of both existing methods to 'create a comprehensive data set that delivers detailed reach and frequency metrics, and audience demographic profiles'.

The company's research in Italy has found that cookie deletion varies by site and by industry. It says that disregarding the impact of cookie deletion and cookie-blocking activities could mean that web site operators are overestimating their audience by as much as 46% in the case of search engines and portals, 43% in telecoms/Internet services, and 23% for news and information sites.

Following trials in Italy, the new Data Integration methodology will be made available as part of the company's Market Intelligence product in Australia early next year. Clients in the US are currently testing the approach.

The conflicts between existing methodologies are clear from the latest search engine rankings from comScore and Hitwise. Both companies state that Google was the top US search engine in July, but their estimates of its share of searches differ. According to Hitwise, Google powered 59.2% of searches, while comScore lists it as taking just a 36.5% share of all searches.

It comes as no surprise to find that the two companies use different methods. comScore's figures are based on a panel of two million consumers using various ISPs, while Hitwise data come directly from ISPs themselves. Heather Hopkins, Director of Research at Hitwise, says that, 'through our partnerships with ISPs we are able to track online behaviour of over 10 million US internet users'. However, comScore questions how representative the sample is. It states that Hitwise gathers data from 'the web logs of an unspecified group of ISPs and toolbar systems. Other than United Online (a free/low cost ISP), no major ISP is known to license its data to another party for such an application, so in the absence of disclosure we can only speculate that smaller regional and other free ISPs form the basis of the sample.'

The companies' web sites are at www.nielsen-netratings.com , www.comscore.com , and www.hitwise.com.

All articles 2006-23 written and edited by Mel Crowther and/or Nick Thomas, 2024- by Nick Thomas, unless otherwise stated.

• Nielsen//NetRatings Australia
• All news