Nielsen ‘regrets’ nine month error which saw Ninemsn video views doubled
Source: Nielsen
Audience measurement company Nielsen says it “regrets” an error which has meant it has been over-reporting the figures for Microsoft and Ninemsn video streams since last July.
Irregularities with March’s figures were first reported in today’s edition of The Australian (paywall), however Nielsen has told Mumbrella the error, which has effectively seen the number of video streams being double counted in the rankings, was going on for nine months.
This is not the first time Nielsen has had a major error in its video streaming numbers and comes at a sensitive time for the company as it prepares to retender to Interactive Advertising Bureau (IAB) preferred provider of online audience measurement contract.
Great article. I have also noticed the ‘9 Newsbreak’ app for IPhones/iPad does not play news clips the first time, it “buffers” and only plays audio. However when it is replayed it works just fine. This applies to every clip and on diffrent iPhones. This suggests they want two hits instead of one, very convenient for doubling the hit counter and viewers.
So, how do we know the video stats for the other publishers are correct? Coming just a few weeks after Neilson released new numbers showing they had “found” an extra 2m internet users in australia, it doesn’t instill a lot of confidence.
I commented about this in July of last year when the questions arose about data released for video streaming at News. I have been waiting for something to come out about it since then. At the time various people stepped forward to defend the system and even though they didn’t really understand my comments had convinced themselves that they could rationalise the numbers being released. It is clear changes to the system were made in the middle of 2013, that had an effect on the data being produced it is unfortunate that it has taken until now to get to the bottom of the issue.
If as much time was spent investigating the issues and fixing them as was spent defending the flawed data may be we would all have been better off.