OzTAM looks to new measurement beyond 2017


OzTAM has invited expressions of interest in tendering for its audience research service contract, ahead of Nielsen’s contract expiry at the end of 2017.

AFR reports the TV ratings provider wrote last week to half a dozen of the world’s biggest research companies including Nielsen, WPP’s Kantar, Australian radio ratings provider GfK, and newspaper readership ratings provider IPSOS.

OzTAM, owned by Seven, Nine and TEN, wants new capabilities for measuring data on mobiles, tablets and laptops and through catch up and video-on-demand services.

“This process will help us determine the approach to delivery of OzTAM’s measurement service from 2018 onwards,” said OzTAM chief executive Doug Peiffer.

The Nielsen panel currently includes 3500 households (covering 8000 people) in Sydney, Melbourne, Brisbane, Adelaide and Perth across all television households and nationally for all subscription television households.


  1. TasTVcameraman

    How about having a truly national system, we in Tasmania do watch TV. I still find it incredible that just 8000 people are our ratings experts.

    Surely with these days of apps for mobile phones, maybe selected people maybe 80000 could be given access to an app, I for one would sign up, and you could have instant results. The app would have to have security protection etc but it could work.

    • Because increasing the sample size tenfold to 80,000 people actually gains you bugger-all in terms of improved accuracy. Very rough “it’s-4:30am-and-I-can’t-sleep” estimate is that you’d need to increase the sample size 40x~50x to get appreciably better than +-1% accuracy.

      And the overnights are too slow for you?

      p.s. at least Tas is represented in RegionalTAM if not OzTAM. The NT isn’t represented in either…

  2. Instead of having a 5-city metro audience, why not make it a 5-state audience e.g. 7/Prime7/GWN7, Nine/WIN/NBN and 10/SC10/Ten West. Metropolitan ratings are different to Regional TAM ratings.

    • Tex, the statistics are only as effective as the sample that is chosen. As far as I know the method of selection (in order to keep the sample as close to the varieties of the population – as OzTam interprets it) is not revealed by OzTam. Therefore any discussion about the ratings accuracy is moot.

      Remember, there are 3 types of lies: lies, damn lies and statistics 😉

      • Which was more or less my point ;).

        FWIW, they do publish enough info publicly to come up with reasonable estimates of accuracy e.g. I think I ballparked “better than 1.7%” here before for the 5-city totals, while independent experts have stated ‘about 1%’.

        (p.s. you don’t want to hear the little aphorisms that statisticians use about people that don’t understand statistics ;))

Leave a Reply

You must be logged in to post a comment.