October 19th, 2014
02:38 PM ET

Should Nielsen's TV ratings be trusted?

Jonathan Klein, a former president of CNN/U.S., speaks with Brian Stelter about the television industry's uncomfortable dependence on Nielsen.

"Ratings rule television. So, really, it's Nielsen that rules television," Stelter said.

Nielsen ratings are what's called the "currency" for the whole TV industry to buy and sell ads and determine what's popular and what's not.

"Nielsen recently admitted to a glitch, a software bug that affected broadcast ratings for months," Stelter said. While the bug was minor, "the incident just deepened the TV industry's mistrust of the company."

Ratings are the "report card" for TV, Klein said, but when he was running CNN, the data never offered enough insight.

"They were a very narrow snapshot of one type of behavior - how many people are sitting in front of a couch watching TV - when, in fact, we know that it's really a panorama. You know, media consumption is about squeezing [TV] in amidst a lot of other things - tweeting, cooking, reading, putting your kids to bed, waking them up, whatever."

"And so, they never felt reliable," Klein said, "and never told us the most important piece of information which is who is watching, and what else are they watching, and what else are they doing, and where else could we find those people, and that's the advantage of digital media over the old obsolete Nielsen ratings."

Nielsen produces viewership estimates for the entire United States population using a scientific sample of 25,000 homes. The company is confident in its sample: "Today," it says, "the ratings are very reliable for the day-to-day, national networks."

The recent glitch affected "a very small amount of total viewing," according to the company.

But when the ratings were corrected, it changed the outcome of the nightly news ratings race - ABC's "World News Tonight" was thought to have beaten the "NBC Nightly News," but it turned out NBC was still #1 among total viewers.

Filed under: Reliable Sources
soundoff (3 Responses)

    Thank you a lot for providing individuals with such a marvellous opportunity to read in detail from this web site. Its usually very excellent plus jam-packed with a lot of fun for me and my office friends to search your website more than 3 times in a week to see the newest things you have got. And of course, Im also certainly astounded with the eye-popping strategies you give. Selected 4 ideas on this page are definitely the most suitable weve had.


    November 26, 2020 at 5:58 am | Reply

    Ive been meaning to read this and just never received a chance. Its an issue that Im extremely interested in, I just started reading and Im glad I did. Youre a good blogger, one of the best that Ive seen. This blog certainly has some info on topic that I just wasnt aware of. Thanks for bringing this things to light.


    November 23, 2020 at 9:30 pm | Reply
  3. John Dickey

    I would like to address something that was said from 2:20 to 2:45 of the show. First, I currently do not work for Nielsen, though I have in the past been in a Nielsen household (maybe 100$ to for 2 years). I would like to address a small concern with your statements about the poll itself. First, a poll of 25,000 homes is an extremely large sample. The size of the sample alone should make it fairly reliable because of the law of large numbers. Moreover we should expect that Nielsen weights its sample on the Census's Current Population Survey and the most recent Population Census to account for changes in the population that are not captured in changes in the Nielsen sample. They can additionally use previous program viewership to either weight samples for current viewership or as a check on the accuracy of their numbers.

    There are some things about the tv industry that could affect the accuracy of the poll. Changing the lineup of programs, the increasing options for programs, and the differences in available programs can have some effect. These are not huge problems and can be accounted for in the data analysis.

    Another thing that came to my mind is the fact that the CNN ORC polls that we hear about in this political season are a substantially smaller sample than that you cited in the program (here you have your company's poll with a sample of 1012 ) . It is still an extremely accurate poll on both the topline number and the crosstabs. Nielsen is a larger poll that has a sample in about 1 house for every 1000 houses. It can account for viewership because so much of the population being examined is in the poll.

    I can understand questioning how they can make an error but questioning the validity of the poll itself seems to be inaccurate. Its a completely other statement to say that its not answering the questions that matter to cable companies. (what other media are people interacting with when the tv is on? what percentage of tv's that are on a station are being watched?) That's a matter of study design.

    October 20, 2014 at 11:41 am | Reply

Post a comment


CNN welcomes a lively and courteous discussion as long as you follow the Rules of Conduct set forth in our Terms of Service. Comments are not pre-screened before they post. You agree that anything you post may be used, along with your name and profile picture, in accordance with our Privacy Policy and the license you have granted pursuant to our Terms of Service.