WEBVTT - How Do TV Ratings Work?

0:00:02.040 --> 0:00:09.760
<v Speaker 1>Welcome to brain Stuff from How Stuff Works. Hey, brain Stuff,

0:00:09.800 --> 0:00:14.000
<v Speaker 1>it's me Christian Seger. We have all heard about TV ratings.

0:00:14.040 --> 0:00:16.640
<v Speaker 1>They're just an estimate of how many people are watching

0:00:16.640 --> 0:00:19.920
<v Speaker 1>a particular show at a given time, and they are

0:00:19.960 --> 0:00:23.640
<v Speaker 1>a big deal. As of people in the US watched

0:00:23.680 --> 0:00:28.200
<v Speaker 1>about thirty four hours of TV each week, Billions of

0:00:28.240 --> 0:00:30.520
<v Speaker 1>AD dollars hang in the balance, and if a show

0:00:30.560 --> 0:00:35.199
<v Speaker 1>doesn't perform, then it risks getting axed. We're all familiar

0:00:35.240 --> 0:00:37.560
<v Speaker 1>with it. More than a few fans have been disappointed

0:00:37.560 --> 0:00:41.839
<v Speaker 1>when low ratings doom their favorite shows to cancelation. And

0:00:41.880 --> 0:00:44.159
<v Speaker 1>without naming any names, it's fair to say that some

0:00:44.240 --> 0:00:48.720
<v Speaker 1>viewers are surprised when shows they hate continue through season

0:00:48.880 --> 0:00:53.400
<v Speaker 1>after season after season due to high ratings. But what

0:00:53.479 --> 0:00:56.639
<v Speaker 1>are these ratings anyways, where do they come from, and

0:00:56.680 --> 0:00:59.680
<v Speaker 1>why are they so important? Well, in the United States

0:00:59.680 --> 0:01:03.880
<v Speaker 1>and Canada, TV ratings are synonymous with one company, Nielsen,

0:01:03.960 --> 0:01:07.040
<v Speaker 1>which was founded in nineteen twenty three by an engineer

0:01:07.120 --> 0:01:12.559
<v Speaker 1>named Arthur Nielsen. Originally, he wanted to sell engineering performance surveys,

0:01:12.640 --> 0:01:16.880
<v Speaker 1>a way to measure the efficiency and quality of engineering operations.

0:01:17.160 --> 0:01:21.679
<v Speaker 1>By ninety two, Nielsen expanded, creating a retail index that

0:01:21.800 --> 0:01:25.200
<v Speaker 1>tracked purchases in the food and drug markets. This was

0:01:25.240 --> 0:01:28.240
<v Speaker 1>the first successful attempt to measure these markets on a

0:01:28.240 --> 0:01:31.959
<v Speaker 1>wide scale, and by nineteen fifty the company applied this

0:01:32.040 --> 0:01:36.960
<v Speaker 1>technique to a little industry called television. Today, Nielsen measures

0:01:37.000 --> 0:01:40.000
<v Speaker 1>the number of people watching television shows and makes its

0:01:40.080 --> 0:01:44.080
<v Speaker 1>data available to cable networks as well as advertisers and

0:01:44.120 --> 0:01:49.120
<v Speaker 1>the media. The company uses a technique called statistical sampling

0:01:49.160 --> 0:01:52.000
<v Speaker 1>to rate the shows. This is the same technique that

0:01:52.080 --> 0:01:56.440
<v Speaker 1>polsters used to predict the outcomes of elections. Nielsen creates

0:01:56.480 --> 0:01:59.240
<v Speaker 1>a sample audience and counts how many people in that

0:01:59.320 --> 0:02:03.040
<v Speaker 1>audience view you each program. They extrapolate from the sample

0:02:03.240 --> 0:02:06.360
<v Speaker 1>and estimate the number of viewers in the entire population

0:02:06.560 --> 0:02:10.000
<v Speaker 1>watching the show to find out who's watching what. The

0:02:10.040 --> 0:02:13.040
<v Speaker 1>company gets thousands of households to become part of the

0:02:13.160 --> 0:02:18.679
<v Speaker 1>representative sample for the national ratings estimates. These participants are

0:02:18.760 --> 0:02:22.280
<v Speaker 1>randomly selected, and they're paid a little bit but not

0:02:22.520 --> 0:02:25.360
<v Speaker 1>near enough to you know, quit their day jobs and

0:02:25.400 --> 0:02:29.280
<v Speaker 1>watch TV full time. Every US household with a TV

0:02:29.520 --> 0:02:32.320
<v Speaker 1>theoretically has a chance to be a part of the sample,

0:02:32.639 --> 0:02:35.720
<v Speaker 1>but the sample itself is not very large. I mean,

0:02:36.040 --> 0:02:41.080
<v Speaker 1>that's just a few thousand households extrapolated to represent millions

0:02:41.200 --> 0:02:46.200
<v Speaker 1>right well. To make up for this, the company measures TVs, homes, programs,

0:02:46.240 --> 0:02:49.160
<v Speaker 1>and people in a variety of ways. The data is

0:02:49.200 --> 0:02:52.639
<v Speaker 1>broken down by demographic, type of stream, and so on.

0:02:53.000 --> 0:02:56.760
<v Speaker 1>This representative sample is compared to the general population, and

0:02:56.880 --> 0:03:00.799
<v Speaker 1>Nielsen also calls thousands of household to see if their

0:03:00.800 --> 0:03:04.080
<v Speaker 1>TV sets are on and who is watching. But the

0:03:04.080 --> 0:03:07.919
<v Speaker 1>phone survey could happen to anyone fitting the criteria, and

0:03:08.080 --> 0:03:10.800
<v Speaker 1>it could also be a one time thing. So what

0:03:10.919 --> 0:03:14.240
<v Speaker 1>about the genuine Nielsen families, you know, the one Nielsen

0:03:14.320 --> 0:03:18.200
<v Speaker 1>monitors continually. Well. To find out what these people are watching,

0:03:18.240 --> 0:03:21.680
<v Speaker 1>the company installs a black box on the TVs in

0:03:21.720 --> 0:03:24.480
<v Speaker 1>a home. This isn't the same as a black box

0:03:24.520 --> 0:03:27.840
<v Speaker 1>on a plane. No, it's just a computer and a modem.

0:03:27.880 --> 0:03:30.600
<v Speaker 1>The box keeps track of when the TV is on

0:03:30.840 --> 0:03:34.240
<v Speaker 1>and what it's tuned to. Every night, the box gathers

0:03:34.320 --> 0:03:37.160
<v Speaker 1>up the households viewing data and sends all of this

0:03:37.280 --> 0:03:41.320
<v Speaker 1>information to the company's central computer. By monitoring what is

0:03:41.360 --> 0:03:44.600
<v Speaker 1>on TV at any given time, the company is able

0:03:44.640 --> 0:03:48.440
<v Speaker 1>to keep track of how many people watch, which program

0:03:48.480 --> 0:03:51.240
<v Speaker 1>that seems fine, but how do we know who is

0:03:51.280 --> 0:03:55.480
<v Speaker 1>watching what? Well, after all, not everyone in a household

0:03:55.600 --> 0:03:58.680
<v Speaker 1>is going to love the same shows. That's where the

0:03:58.840 --> 0:04:02.440
<v Speaker 1>people meters come in. These are small boxes placed near

0:04:02.480 --> 0:04:05.520
<v Speaker 1>the TV sets of those in the national sample. They

0:04:05.560 --> 0:04:08.200
<v Speaker 1>measure who is watching by giving each member of the

0:04:08.240 --> 0:04:11.200
<v Speaker 1>household a button to turn on and off to show

0:04:11.240 --> 0:04:15.520
<v Speaker 1>when he or she begins and ends viewing. This information

0:04:15.600 --> 0:04:19.520
<v Speaker 1>is also collected each night. The national TV ratings have

0:04:19.600 --> 0:04:24.200
<v Speaker 1>relied on these meters for years. To ensure reasonably accurate results,

0:04:24.480 --> 0:04:28.840
<v Speaker 1>the company uses audits and quality checks. They also regularly

0:04:29.120 --> 0:04:33.640
<v Speaker 1>compare the ratings they get from different samples and measurement methods. So,

0:04:33.880 --> 0:04:37.880
<v Speaker 1>for example, a one point oh Nielsen rating indicates that

0:04:37.960 --> 0:04:40.560
<v Speaker 1>one percent of the one hundred and fifteen point nine

0:04:40.640 --> 0:04:45.719
<v Speaker 1>million estimated TV watching households tuned into a program. The

0:04:45.839 --> 0:04:49.240
<v Speaker 1>data is also broken off into different demographic ratings, the

0:04:49.279 --> 0:04:53.040
<v Speaker 1>most important being people ages eighteen to thirty four. Now,

0:04:53.080 --> 0:04:57.120
<v Speaker 1>make no mistake, this research is worth billions of dollars.

0:04:57.320 --> 0:05:00.760
<v Speaker 1>Advertising rates are based on Nielsen's data. That's why a

0:05:00.839 --> 0:05:04.120
<v Speaker 1>thirty second commercial on one show might cost twice as

0:05:04.200 --> 0:05:07.560
<v Speaker 1>much as a commercial on a low rated show programmers

0:05:07.600 --> 0:05:11.239
<v Speaker 1>also use Nielsen's data to decide which shows to keep

0:05:11.360 --> 0:05:15.000
<v Speaker 1>and which to cancel. A show that has several million

0:05:15.080 --> 0:05:18.640
<v Speaker 1>viewers may seem popular to us, but a network may

0:05:18.640 --> 0:05:21.640
<v Speaker 1>need millions more watching that program to make it a

0:05:21.680 --> 0:05:26.400
<v Speaker 1>financial success. That's why some shows with loyal following still

0:05:26.440 --> 0:05:30.240
<v Speaker 1>get canceled. Sorry, Firefly, and there's an elephant in the

0:05:30.320 --> 0:05:34.200
<v Speaker 1>room here too. The way people watch TV is changing.

0:05:34.440 --> 0:05:38.600
<v Speaker 1>With DVRs, Netflix and other streaming services, TV viewers are

0:05:38.680 --> 0:05:42.880
<v Speaker 1>more likely to customize their viewing habits, watching stuff when

0:05:42.920 --> 0:05:45.279
<v Speaker 1>they want to see it, rather than when it happens

0:05:45.320 --> 0:05:48.640
<v Speaker 1>to be on Nielsen has ways of measuring some of this,

0:05:48.839 --> 0:05:52.039
<v Speaker 1>but not all of it. As viewing habits continue to

0:05:52.120 --> 0:05:57.920
<v Speaker 1>fragment across different platforms, advertisers, content creators, and audience members

0:05:57.920 --> 0:06:02.279
<v Speaker 1>alike are right to ask how accurate these ratings actually are.

0:06:07.240 --> 0:06:09.480
<v Speaker 1>Check out the brainstuff channel on YouTube, and for more

0:06:09.520 --> 0:06:12.080
<v Speaker 1>on this and thousands of other topics, visit how stuff

0:06:12.080 --> 0:06:26.680
<v Speaker 1>works dot com.