Fraternal twins
Significant Differences Between ACNielsen and MMXI Internet Panels
Major differences between ACNielsen and MMXI Internet panels by Markus KnöpfliACNielsen has published the data from its Internet home panel for Switzerland for the first time. However, although the panel is almost identical to that of MMXI Switzerland, the results differ considerably in some cases. The usage time in April alone shows a difference of over half an hour.
According to data published for the first time by Nielsen/Netratings, 1.57 million people in Switzerland used the web at home for an average of six hours in April. MMXI Switzerland reports completely different data for April (WW 21/01): According to this, 2.195 million people used the web for around five and a half hours.
The differences are surprising, as the panels are almost identical: Although MMXI clearly has the larger panel with 4863 people, ACNielsen falls slightly short with 3333 participants. MMXI estimates the population of people with an Internet connection at home at 2.8 million, ACNielsen at 2.99 million. Both carry out monthly surveys in order to adjust the data on an ongoing basis. And both measure Internet use on the panel participants' PCs using software that transmits the data simultaneously to a central office.
And yet the biggest difference is already in the extrapolated number of active Internet users, which differs by 600,000 people (!). A second significant difference lies in the ratings. The top 20 domain list from Nielsen/Netratings only resembles that of MMXI in the first nine places, while there are striking differences further down the list. In 10th and 11th place, ACNielsen lists Geocities.com and Bluewindow.ch, two candidates that are not in the top 20 of MMXI. And in 13th, 17th and
19, Nielsen/Netratings brings further "outsiders" into play with Netscape.com, Gmx.net and Yahoo.fr. Others such as Altavista.com or Swissonline.ch make great leaps forward or backward, depending on the panel.
Attempts at explanation are not beyond all doubt
There are attempts to explain the differences on both sides, but they are not very credible. ACNielsen claimed that MMXI was primarily targeting heavy users when recruiting panel participants, while Nielsen/Netratings only asked whether an Internet connection was available at home.
However, this is disputed by MMXI: they also ask for an Internet connection, they say. MMXI complains that the Nielsen/Netratings software does not recognize an interim switch to offline applications, which is why higher usage values are shown. However, the difference of half an hour seems too small for this.
However, the ACNielsen panel has an advantage that MMXI cannot (yet) offer: It measures what percentage of net visitors click on a banner or button. The average click rate in April was 0.43 percent. New forms of web advertising are constantly being integrated into the measurement software.
According to data published for the first time by Nielsen/Netratings, 1.57 million people in Switzerland used the web at home for an average of six hours in April. MMXI Switzerland reports completely different data for April (WW 21/01): According to this, 2.195 million people used the web for around five and a half hours.
The differences are surprising, as the panels are almost identical: Although MMXI clearly has the larger panel with 4863 people, ACNielsen falls slightly short with 3333 participants. MMXI estimates the population of people with an Internet connection at home at 2.8 million, ACNielsen at 2.99 million. Both carry out monthly surveys in order to adjust the data on an ongoing basis. And both measure Internet use on the panel participants' PCs using software that transmits the data simultaneously to a central office.
And yet the biggest difference is already in the extrapolated number of active Internet users, which differs by 600,000 people (!). A second significant difference lies in the ratings. The top 20 domain list from Nielsen/Netratings only resembles that of MMXI in the first nine places, while there are striking differences further down the list. In 10th and 11th place, ACNielsen lists Geocities.com and Bluewindow.ch, two candidates that are not in the top 20 of MMXI. And in 13th, 17th and
19, Nielsen/Netratings brings further "outsiders" into play with Netscape.com, Gmx.net and Yahoo.fr. Others such as Altavista.com or Swissonline.ch make great leaps forward or backward, depending on the panel.
Attempts at explanation are not beyond all doubt
There are attempts to explain the differences on both sides, but they are not very credible. ACNielsen claimed that MMXI was primarily targeting heavy users when recruiting panel participants, while Nielsen/Netratings only asked whether an Internet connection was available at home.
However, this is disputed by MMXI: they also ask for an Internet connection, they say. MMXI complains that the Nielsen/Netratings software does not recognize an interim switch to offline applications, which is why higher usage values are shown. However, the difference of half an hour seems too small for this.
However, the ACNielsen panel has an advantage that MMXI cannot (yet) offer: It measures what percentage of net visitors click on a banner or button. The average click rate in April was 0.43 percent. New forms of web advertising are constantly being integrated into the measurement software.
