Nielsen says digital news rankings for April were wrong, and Guardian didn’t beat Daily Mail after all
Nielsen has reissued its news website rankings for last month, saying that an error led it to incorrectly indicate that The Guardian had overtaken Daily Mail for Australian audience.
All news sites in the top 10, except The Guardian and Yahoo7 News websites were said to have experienced a decline over the month, however Nielsen has today admitted a “data production issue” led to under-reporting. The Daily Mail Australia actually only slipped 8% and retained its top five position, ahead of The Guardian.
News.com.au, which was said to have declined 2% to 5.630m, actually climbed 5% to 6.015m. ABC News websites climbed 1% to 4.912m, in contrast to the previously reported decline of 5% over the month. Nine.com.au climbed 3% to 4.698m, rather than the 3% decline Nielsen initially cited, and SMH.com.au dropped 1% to 4.078m, not 7% as previously reported.
In reissuing the data one week later, Nielsen noted smartphone and tablet data had not been collected correctly.
“The re-statement is due to a data production issue that resulted in smartphone and tablet data for some websites to be under-reported for the month of April 2017. Originally, Nielsen had attributed the declines from March 2017 to April 2017 to the occurence of Easter, ANZAC Day and school holidays. An in-depth analysis was conducted and uncovered the production issue,” a statement said.
Nielsen’s executive director of media Monique Perry further explained the problem, saying: “Nielsen found smartphone and tablet audiences were under-reported for entities containing URLs which were first observed in the data during a 10-day period from April 20 to April 30. The average unique audience impact to all 3,285 reported brands was approximately 3%. Of this, 287 brands were impacted by more than 10%.”
Perry noted once all URLs were correctly included, all impacted brands saw an increase in the figures which were originally released last week, but Nielsen said the changes were minor.
“There was very little impact to the total market line with no change to total online audience. Sessions per person increased by 1% and time per person by 2%,” the statement said.
Nielsen said it had now implemented additional automated and manual checks to “proactively identify and correct any issues in the future”, which would “ensure this issue cannot happen again”.
The re-issued figures
Daily Mail Australia managing director Peter Holder said the publication had noticed the error and was pleased it has been rectified.
“Last week’s result was a bit like getting a phone bill and knowing you’re being slugged for calls you didn’t make. It just didn’t make sense but if you say nothing, nothing will be done and you’ve only got yourself to blame,” Holder said.
“We simply said to Nielsen – a few times admittedly – that our data and other methodologies were showing a single digit percentage drop that was seasonally consistent: April always being affected by school holidays, Easter and Anzac Day; coupled with the fact it was a relatively quiet month for big news stories.
“We’re delighted Nielsen listened to our reasoning and responded in the way in which it did.”
topics
digital publishing
Monique Perry
Nielsen
Nielsen Digital Ratings (Monthly)
nielsen online news rankings
online news
Why does anyone use this tool for measurement? Nielsen have no idea what they are doing. It’s actually embarrassing.
User ID not verified.
Well I suppose if it’s good enough for Facebook and YouTube to make super public cockups, Nielsen might as well join the club.
User ID not verified.
Damn. A small flicker of hope for humanity extinguished right there…
User ID not verified.
Nielsen, about as accurate as horoscopes
User ID not verified.
The digital advertising industry gets more comical by the day, you couldn’t make this stuff up….No pun intended.
User ID not verified.
Well at least there is tracking ability as opposed to assumptive hogwash with traditional. ‘Readership’ being one of a variety of examples.
User ID not verified.
Sure ain’t perfect – but beats the shit out of vendor data, not the least because it reports audience instead of browsers, page views etc.
Anyone else remember when the industry were quite happy to report 130 million ‘Unique Browers’ in Australia?
User ID not verified.
Great Article! Thanks for taking the time to put this together.
User ID not verified.