Cable TV Viewership: More Viewing Declines In May, Streamers Stronger: Nielsen
- by Wayne Friedman , June 22, 2023
Although the NBA playoffs on ESPN and TNT boosted sports viewership in May, overall cable TV viewing continues to decline -- down 14% in the most recent reading of Nielsen’s The Gauge.
The Gauge, which measures total day viewership of persons two years and older, now gives total cable TV viewing a 31.1 share -- versus a 36.5 share in the year-ago monthly period.
This was largely attributable to cable TV news viewing -- a major component of cable TV viewing. Nielsen says this was down 11% from the previous month.
Broadcast viewing also declined year-over-year -- down 5.6% to a 22.8% share.
After a slight dip in April, streaming usage continues its march to higher ground, now at a dominant 36.4% share -- up from 31.9% a year ago. Major gains were seen with Netflix -- up 9.2% (to a 7.9% share) -- and Amazon Prime Video, 5.1% higher to a 3.1% share.
advertisement
advertisement
Nielsen says it has made a change in methodology and now credits viewing of streaming originals, through cable set-top boxes, in the streaming category. Viewing was previously included in the “Other” category.
At the same time, Nielsen says YouTube has been growing, and is now at a 8.5% share, versus 6.7% a year ago, while Hulu is at 3.7%, up from 3.4% a year ago.
Nielsen notes a year ago YouTube and Hulu respective linear virtual pay TV businesses were included as part of the streaming category. In February of this year, those businesses were excluded from the streaming category.
Nielsen was basically double-counting viewing. Ther were counting viewing from vMVPD's and counting it as streaming (which they should) and as cable & broadcast (which they should not). To begin with, that's complete nonsense. In either March or April Nielsen decided to take any VMVPD viewing and only count it as cable and broadcast viewing. Even though it is not taking place on broadcast or cable they are giving 100% credit and reducing the streaming number, more nonsense. That's what I expect from Nielsen.
Does anyone know the sample size of Gauge? I have been trying to find that out. When it was first introduced it was 14,000 to measure the entire US all Broadcast, cable, and hundreds of apps. Sorry, just not a fan of nielsen. I am thankful that I can combine Nielsen's numbers with several other reputable research companies.
Robert, I believe that Nielsen melds togther data from two panels----one which has the ability to monitor linear plus streaming and the other which is only for linear. But sample size is not the issue when tabulating so much viewing----- which amounts to 4.5-5 hours per day for an average panel member. Anything over 3,000 or so would be OK for such a purpose and Nielsen's sample is far larger than that ---though I can't give you the exact current figure.
The real question should be are these panels representative of the nation as a whole in terms of their demographic composition, the kinds of sets they have, etc.? So far I have yet to see any evidence that Nielsen's panels are way off the mark regarding national projectability---so if we define "streaming" as any activity that takes place on a TV set and Nielsen meters all of the sets per home using a reasonably sized panel that is reasonably representative, the numbers should be considered to be reasonably accurate. Not to be tken literally---no survey offers perfect accuracy---but not to be dismissed simply because of other issues---like Nielsen's stumble during the early days of Covid-19 or its past problems with the MRC.