Time-Shifted Viewing Up, Live Viewing Dips

by Wayne Friedman , September 11, 2012

WatchingTV-ABC Live television viewing slipped again in the first quarter of this year, according to Nielsen.

The average daily live TV viewing per person is now at 4 hours and 38 minutes -- down from 4:47 during the first quarter of 2011. Time-shifted viewing climbed to 24 minutes from 21 minutes. Gaming time stayed the same at 14 minutes a day, with DVD playback slipping to an average 12 minutes from 14 minutes.

Nielsen Cross-Media report says on a weekly basis, U.S. viewers are watching nearly 35 hours per week of video across all screens, and spending five hours a week using the Internet. Gaming time now is one hour and 38 minutes per week on average. But among those who watch streaming video content on gaming devices, it is now 2 hours and 48 minutes.

Nielsen says traditional TV viewing is slipping but shifting to other devices -- especially tablets. Currently, more than 15% of U.S. TV homes have one. Also factoring into all this media change -- nearly 36 million mobile phone owners in the U.S. watch video on their phones.

TV home penetration is at nearly 96% and slowly declining. But Nielsen says "in homes that are not considered traditional TV homes," 75% own a TV set -- probably connected to a gaming device, over-the-top device, DVD player or the Internet directly.

cross-platform, media measurement, tv

3 comments about "Time-Shifted Viewing Up, Live Viewing Dips".

Check to receive email when comments are posted.

Doug Garnett from Protonik, LLC, September 11, 2012 at 4:41 p.m.
Wayne - Does Nielsen deal with the impact of the recession in the TV home penetration questions? The public reporting is pretty cagey about how they come up with their penetration numbers.
Reply

Doug Garnett from Protonik, LLC, September 11, 2012 at 4:46 p.m.

Saw another report - Nielsen's not clear how much decline is real and how much is because of the recession. I'd also be interested to know how the "no longer a TV HH" homes rank on hours per week of TV. Seems a reasonable guess that they're the low-volume TV consumers.

John Grono from GAP Research, September 12, 2012 at 12:33 a.m.

I'm not 100% certain how the US do it, but I suspect it is pretty much the same as Australia. We use Census data to construct a sampling frame from which we conduct an 'Establishment Survey' - generally 10 times the ratings sample so probably in the order of hundreds of thousands of homes. In Australia we contact homes via landline and mobile in representative proportions. We then collect information that official bodies don't collect that is correlated to TV viewing - size of the household, working status, number of TVS, number for PVR devices, HD penetration, internet connection (and speed), games and console ownership, tablet and mobile ownership, cable subscriber etc. ... and of course how many homes don't own a working TV as well!

From that Establishment Survey you can first quantify the 'universe' number by those parameters, and second recruit a sample that represents the composition of the market.

Therefore, if there is a decline in TV ownership, cable subscription etc, the sample composition and markect size that the sample is projected to also declines.

Basically, a "no longer a TV home" is catered for by taking the US population and removing the percentage found during the Establishment Survey who don't own a working TV. So if there was (say) 110 million homes and 5% in the Establishment Survey didn't have a working TV, you would set the TV univers at 110 * 0.95 = 104.5 million homes.

A "no TV home" racks up zero hours a week, unless you are also monitoring computer, mobile and tablet devices. So far, we're only monitoring computers in some of the panel homes here and finding comparatively low usage levels.